<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Rishika Gupta on Medium]]></title>
        <description><![CDATA[Stories by Rishika Gupta on Medium]]></description>
        <link>https://medium.com/@rishikagupta1999?source=rss-6521c7f86bab------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*8lNzziXTn8WKGJ00r3n37Q.jpeg</url>
            <title>Stories by Rishika Gupta on Medium</title>
            <link>https://medium.com/@rishikagupta1999?source=rss-6521c7f86bab------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 05:10:41 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@rishikagupta1999/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Interview Experience L’Oréal — Data Science and Machine Learning Internship]]></title>
            <link>https://rishikagupta1999.medium.com/interview-experience-lor%C3%A9al-data-science-and-machine-learning-internship-eb4f27814626?source=rss-6521c7f86bab------2</link>
            <guid isPermaLink="false">https://medium.com/p/eb4f27814626</guid>
            <category><![CDATA[interview]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[data-scientist]]></category>
            <category><![CDATA[interview-questions]]></category>
            <category><![CDATA[data-analysis]]></category>
            <dc:creator><![CDATA[Rishika Gupta]]></dc:creator>
            <pubDate>Tue, 16 Sep 2025 19:49:16 GMT</pubDate>
            <atom:updated>2025-09-16T19:51:52.528Z</atom:updated>
            <content:encoded><![CDATA[<h3>Interview Experience L’Oréal — Data Science and Machine Learning Internship</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4OccDWgWXQcDVPAxWQPePg.png" /><figcaption>L’Oréal Groupe</figcaption></figure><p>Embarking on a journey with L’Oréal was more than just a series of interviews — it was an experience that spanned months, culminating in a tapestry of moments, each holding its significance. From the initial application on a summer day to the final engaging conversation with the Data and Insights Team Manager, every step was a chapter in a story of pursuit and perseverance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Kzwam_EF4wV6YvKh" /><figcaption>Photo by <a href="https://unsplash.com/@mkvandergriff?utm_source=medium&amp;utm_medium=referral">Maranda Vandergriff</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>In this article, I’ll walk you through my immersive interview process with L’Oréal, sharing insights, reflections, and invaluable lessons learned along the way.</p><p>I applied for the position of Data Science and Machine Learning Engineer. Following the initial resume screening, I proceeded to my first interview, conducted by the Talent Acquisition team specialising in Data for Beauty Tech. Explore the subsequent section to delve deeper into the interview, unveiling the questions and insights shared during this phase.</p><h4>Q) Could you introduce yourself?</h4><p>Starting with my introduction, wherein I talked about my current education, along with my foundations relevant to the role. To conclude, I spoke about what excites me the most at L’Oréal and my enthusiasm to work at the firm.</p><h4>Q) What is one of the projects that I am most proud of?</h4><p>I started talking about one of the projects I worked on in my previous role at a company. Following the STAR format — Situation, Task, Action and Result, I spoke about the significant issue I encountered and the task I took to resolve the issue. I highlighted my actions that showed team collaboration, understanding and technical abilities. For the result, I mentioned how my actions impacted my team’s productivity and how this will help me tackle similar challenges at L’Oréal.</p><h4>Follow-up Q) What was the most challenging task for you in that project?</h4><p>Referring to my previous answer, I mentioned how collaboration and understanding the technicalities at each step were essential for my task. And further pressing on how such a simple thing can be challenging.</p><h4>Q) Why did you want to work with L’Oréal?</h4><p>To answer this, I elaborated on what the most genuine things are that excite me about the opportunity of working at L’Oréal. Some of the key ideas that I highlighted in this answer were —</p><ul><li>L’Oréal has the power of technology and innovation to create beauty that can change the world, and how perfectly it aligns with the values or ideas that I want to accomplish in my life.</li><li>Next, the beauty industry is continuously evolving, so I see immense potential in the field of data science and machine learning to drive innovation in this space.</li><li>I also talked about the benefits — connect and learn platform, working with talented peers and events that will help me grow professionally as well as personally.</li></ul><h4>Q) What do you understand by beauty? Or what does it mean to you?</h4><p>Beauty, for me, is about being yourself and feeling confident. It’s about celebrating what makes each person unique and using innovation to include everyone. It’s all about feeling good and empowered in your own self.</p><h4>Q) Am I interviewing with other companies or institutions?</h4><p>So, let me be honest, at the moment I was not interviewing with any other company, so I said no — L’Oréal is the only one. But I will need to look for more opportunities with other companies for my internship.</p><p>Immediately after I finished, I realised that it was such a stupid answer. Instead, I should have elaborated that I am actively exploring opportunities for my thesis project with a few different companies. I believe that this approach allows me to explore a diverse range of projects and industries to find the best fit for both my academic goals and the company’s needs. However, I want to emphasise that my interest in L’Oréal is genuine, and I’m excited about the possibility of contributing my skills and expertise to your organisation. My priority is to find a meaningful and mutually beneficial partnership that aligns with my academic and career objectives.</p><h4>Q) Given a magic wand, what would be the role I would want to work in?</h4><p>I explained how I wanted to work as a data analyst/scientist and also how the tasks accomplished in this role can help the company and its customers. PS — This question is just to understand what kind of role I am looking for.</p><h4>Follow-up Q) How will you explain the difference between the roles of data analyst and data scientist to a non-technical teammate?</h4><p>Imagine a data analyst as a detective with a magnifying glass. Their job is to examine data that a company already has. They look for patterns, trends, and insights within this data to answer specific questions or solve existing problems. For example, they might analyse sales data to figure out which products are selling the most in a particular region.</p><p>Now, picture a data scientist as a wizard. They not only have a magnifying glass, but also a crystal ball! Data scientists not only explore existing data but also predict future outcomes. They create magical models and use advanced tools to foresee what might happen based on past data. For instance, they could predict future sales based on previous sales patterns and other influencing factors.</p><p>In simple terms, a data analyst focuses on understanding what has happened or is happening, like a detective analysing clues. On the other hand, a data scientist takes it a step further, predicting what could happen in the future, like a wizard predicting the outcome of a magical event based on known patterns and rules. Both roles are vital for a company to make informed decisions and steer in the right direction.</p><h4>Q) If you have 3 or more companies to choose an offer from, what will be your top 3 parameters?</h4><p>The interviewer said, and I quote — “There is no right or wrong answer to this question”. Well, I answered that:</p><ul><li>How do the company’s missions and objectives align with my values and principles?</li><li>How does the job description fit my experience and requirements?</li><li>How well is the company committed to ensuring diversity and inclusion principles?</li></ul><p>Some other points I had in mind were —</p><ul><li>The company’s strong commitment to sustainability issues and protecting the environment</li><li>Growth potential at the company</li></ul><p>Since the interview was already nearing the end, I was asked if I had any questions for the interviewer, to which I asked — <strong>What is a typical career progression for someone joining this role?</strong></p><p>Following this stage, I engaged in an interview with my prospective manager. It resembled more of an in-depth discussion, aimed at understanding the team’s responsibilities, aligning expectations, and exploring how my potential role could seamlessly integrate within the existing team dynamics. Following this dialogue, the official offer was extended, and I received confirmation of my selection for the position.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*w9T0hyGKxcDcnA1z" /><figcaption>Photo by <a href="https://unsplash.com/@yulokchan?utm_source=medium&amp;utm_medium=referral">Joseph Chan</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>As I eagerly anticipate my Master’s Internship with L’Oréal, I reflect on the invaluable lessons learned and insights gained during this immersive journey. From discussing technical projects to understanding the company’s ethos, every step unveiled a new facet of the organisation and its dedication to innovation, diversity, and creating a positive impact. Thank you to L’Oréal for this incredible opportunity and to those who supported me along this path.</p><p><em>Thanks for reading! Do subscribe and share with everyone in your community and don’t forget to like and comment. Follow my Medium Page </em><a href="https://medium.com/u/6521c7f86bab"><em>Rishika Gupta</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=eb4f27814626" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ The Ultimate Git Cheatsheet for Developers and Data Engineers]]></title>
            <link>https://python.plainenglish.io/the-ultimate-git-cheatsheet-for-developers-and-data-engineers-f3a2cb39cf43?source=rss-6521c7f86bab------2</link>
            <guid isPermaLink="false">https://medium.com/p/f3a2cb39cf43</guid>
            <category><![CDATA[data-engineering]]></category>
            <category><![CDATA[cheatsheet]]></category>
            <category><![CDATA[git]]></category>
            <category><![CDATA[github]]></category>
            <category><![CDATA[developer]]></category>
            <dc:creator><![CDATA[Rishika Gupta]]></dc:creator>
            <pubDate>Sun, 08 Jun 2025 15:09:28 GMT</pubDate>
            <atom:updated>2025-06-11T00:13:46.589Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*bJXb3EIp1xHZgwhP" /><figcaption>Photo by <a href="https://unsplash.com/@synkevych?utm_source=medium&amp;utm_medium=referral">Roman Synkevych</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Whether you’re a data engineer pushing production pipelines, a developer collaborating on open source, or a student learning version control, Git is an essential skill. In this guide, I’ll walk you through the most commonly used Git commands with clear explanations and practical examples.</p><p>Let’s dive in.</p><h3>1. 🔧 Basic Git Setup</h3><p>Before you start using Git, configure your identity. Setting up your identity ensures that all your commits are correctly attributed to you. This is critical for collaboration, code ownership, and accountability, especially in team or open-source environments.</p><pre>git config --global user.name &quot;Jane Doe&quot;<br>git config --global user.email &quot;jane.doe@example.com&quot;</pre><p>💡 Tip: Use --global to apply settings globally. Omit it for per-project settings.</p><h3>2. 📁 Repository Basics</h3><p>Every project starts with a repository as they are the foundation of every Git project. Whether starting from scratch or collaborating on existing projects, these commands help you set up or join development seamlessly.</p><pre>git init</pre><p>Creates a new local Git repository in the current directory.</p><pre>git clone https://github.com/username/repo.git</pre><p>Clones a remote repository to your local machine.</p><h3>3. 🌿 Working with Branches</h3><p>Branching is one of Git’s core features. It is incredibly useful, allowing multiple people to work on the same repository (parallel development &amp; experimentation). Everyone can work independently on features or fixes without interfering with each other. Eventually, all the changes can be merged into a main branch.</p><pre>git branch</pre><p>Lists all branches in your repository.</p><pre>git branch feature/data-cleanup</pre><p>Creates a new branch — data-cleanup.</p><pre>git checkout feature/data-cleanup</pre><p>Switches to the branch — data-cleanup.</p><pre>git checkout -b feature/api-integration</pre><p>Creates <strong>and</strong> switches to a new branch — api-integrationin one step.</p><pre>git merge feature/api-integration</pre><p>Merges another branch into your current one.</p><h3>4. ✍️ Making and Staging Changes</h3><p>Here’s how you track what’s changed and prepare commits. Staging allows you to control exactly what goes into a commit, and committing helps you capture milestones in your development. Think of this like versioned snapshots — it’s key to tracking progress and debugging.</p><pre>git status</pre><p>Shows modified, new, and deleted files.</p><pre>git add script.py</pre><p>Stages a specific file, in this case — script.py</p><pre>git add .</pre><p>Stages all changes in the current directory.</p><pre>git commit -m &quot;Describe the change&quot;</pre><p>Commits staged changes with a meaningful message.</p><h3>5. 👀 Tracking and Viewing Changes</h3><p>Compare your working directory, staged changes, and commit history. These commands let you audit your codebase. They are essential for reviewing work, troubleshooting bugs, and understanding what has changed over time, especially helpful in team reviews and during debugging sessions.</p><pre>git diff</pre><p>Shows unstaged changes.</p><pre>git diff --staged</pre><p>Shows differences between staged files and the last commit.</p><pre>git log</pre><p>Shows a chronological list of commits.</p><h3>6. 🌐 Pushing and Pulling Changes</h3><p>Time to sync with your remote repository (e.g., GitHub, GitLab) as your work doesn’t live in isolation. These commands are the bridge between your local changes and the shared team repository. They ensure everyone stays up-to-date and reduces merge conflicts.</p><pre>git push origin main</pre><p>Pushes your changes to the main branch on the remote.</p><pre>git pull origin main</pre><p>Fetches and merges changes from the remote main branch.</p><h3>7. 🧹 Undoing Changes (Carefully!)</h3><p>Made a mistake? Git’s got your back. Git provides flexible ways to undo changes — from resetting a single file to reverting an entire commit — so you can experiment fearlessly and recover quickly.</p><pre>git checkout -- data.csv</pre><p>Reverts changes in the working directory (⚠️ destructive), particularly the file — data.csv.</p><pre>git reset script.py</pre><p>Unstages script.py,but keeps changes.</p><pre>git reset --hard a1b2c3d</pre><p>⚠️ Hard reset to a previous commit — a1b2c3d. This will <strong>delete all changes</strong> since then.</p><pre>git revert a1b2c3d</pre><p>Safely undoes a commit by creating a new “reverting” commit to — a1b2c3d.</p><h3>8. 🏷️ Tagging and Releases</h3><p>Tags help mark release points like v1.0, making it easy to refer to stable versions of your code. They&#39;re particularly important in CI/CD pipelines and version management.</p><pre>git tag v1.0</pre><p>Creates a local tag of v1.0</p><pre>git push origin v1.0</pre><p>Pushes the tag — v1.0, to the remote repository.</p><h3>9. 🧠 Advanced Git Commands (Perfect for Data Engineers)</h3><p>As a data engineer or backend developer, you often work on large, interdependent systems. These advanced commands help you handle temporary changes, bring in hotfixes, or rewrite history for cleaner project logs. In simpler terms, these commands help manage messy work, isolate bug fixes, or integrate changes smoothly.</p><pre>git stash</pre><p>Temporarily saves your uncommitted changes.</p><pre>git stash apply</pre><p>Reapplies the stashed changes.</p><pre>git cherry-pick 3e4f1d2</pre><p>Applies a specific commit from another branch — 3e4f1d2.</p><pre>git rebase main</pre><p>Reapplies commits from your branch on top of main, creating a cleaner commit history.</p><h3>✅ Final Thoughts</h3><p>Git can feel overwhelming at first, but with a cheatsheet like this, you’ll gain confidence quickly. Bookmark this guide or clone <a href="https://training.github.com/downloads/github-git-cheat-sheet.pdf">GitHub’s official cheat sheet</a> for reference.</p><p>If you’re a data engineer juggling scripts, notebooks, and DAGs — learning Git properly is an investment worth making.</p><p><em>💬 Have a favorite Git trick or command I missed? Drop it in the comments!</em></p><p><em>Thanks for reading! Do subscribe and share with everyone in your community, and don’t forget to like and comment. Follow my Medium Page, </em><a href="https://medium.com/u/6521c7f86bab"><em>Rishika Gupta</em></a><em>.</em></p><h3>Thank you for being a part of the community</h3><p><em>Before you go:</em></p><ul><li>Be sure to <strong>clap</strong> and <strong>follow</strong> the writer ️👏<strong>️️</strong></li><li>Follow us: <a href="https://x.com/inPlainEngHQ"><strong>X</strong></a> | <a href="https://www.linkedin.com/company/inplainenglish/"><strong>LinkedIn</strong></a> | <a href="https://www.youtube.com/@InPlainEnglish"><strong>YouTube</strong></a> | <a href="https://newsletter.plainenglish.io/"><strong>Newsletter</strong></a> | <a href="https://open.spotify.com/show/7qxylRWKhvZwMz2WuEoua0"><strong>Podcast</strong></a> | <a href="https://twitch.tv/inplainenglish"><strong>Twitch</strong></a></li><li><a href="https://differ.blog/"><strong>Start your own free AI-powered blog on Differ</strong></a> 🚀</li><li><a href="https://discord.gg/in-plain-english-709094664682340443"><strong>Join our content creators community on Discord</strong></a> 🧑🏻‍💻</li><li>For more content, visit <a href="https://plainenglish.io/"><strong>plainenglish.io</strong></a> + <a href="https://stackademic.com/"><strong>stackademic.com</strong></a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f3a2cb39cf43" width="1" height="1" alt=""><hr><p><a href="https://python.plainenglish.io/the-ultimate-git-cheatsheet-for-developers-and-data-engineers-f3a2cb39cf43">🚀 The Ultimate Git Cheatsheet for Developers and Data Engineers</a> was originally published in <a href="https://python.plainenglish.io/">Python in Plain English</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Feature Scaling]]></title>
            <link>https://python.plainenglish.io/feature-scaling-c16f58f1f017?source=rss-6521c7f86bab------2</link>
            <guid isPermaLink="false">https://medium.com/p/c16f58f1f017</guid>
            <category><![CDATA[scaling]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[data-engineering]]></category>
            <category><![CDATA[feature-engineering]]></category>
            <category><![CDATA[computer-science]]></category>
            <dc:creator><![CDATA[Rishika Gupta]]></dc:creator>
            <pubDate>Sun, 28 Jan 2024 22:04:15 GMT</pubDate>
            <atom:updated>2024-01-29T06:25:59.486Z</atom:updated>
            <content:encoded><![CDATA[<h3>Feature Scaling in Python</h3><figure><img alt="Feature scaling" src="https://cdn-images-1.medium.com/max/1024/1*z7CtD6-C5KTKP7Q8hZnNsg.png" /><figcaption>Image reference: <a href="https://editor.analyticsvidhya.com/uploads/41951FEATURE%20SCALING%20IN%20PYTHON.png">Analytics Vidya</a></figcaption></figure><p>In this article, we will discuss the concept of feature scaling, why is it required and some basic models used in practice in Python.</p><h4>1. Introduction to Feature Scaling</h4><p>In the realm of scientific endeavours, the concept of scaling entails adjusting the range of variables to enhance clarity or accommodate specific needs. When delving into machine learning, scaling primarily refers to the adjustment of feature ranges during data preprocessing before training models with machine learning algorithms.</p><p>Consider a dataset featuring disparate ranges for different features, such as ages distributed between 0 and 80 years and salaries spanning 0 to 1 million euros. Without proper scaling, a linear regression algorithm might assign disproportionate weight to features with larger ranges, potentially skewing the model’s performance.</p><p>Feature scaling comes to the rescue in such scenarios. By standardizing the range of independent variables or features, typically within a specified range like [-1, 1] or [0, 1], the goal is to ensure equal importance for each feature. This not only prevents a single feature from dominating others but also facilitates quicker and more effective training and tuning of machine learning algorithms.</p><p>It’s worth noting that various methods, both linear and non-linear, can be employed for feature scaling. Options range from min-max scaling and z-score standardization to techniques like clipping, winsorizing, or even taking the logarithm of inputs before scaling. The choice of method hinges on the specific characteristics of your data and the machine learning algorithm at play.</p><h4>2. Why do care about this?</h4><p>Machine learning algorithms, utilize optimization or learning techniques like gradient descent optimizers optimized for the [-1, 1] numeric range, necessitating the scaling of numeric values to conform to this range. The primary rationale behind this requirement goes as follows:</p><ul><li>Optimization algorithms, such as gradient descent, demand more iterations to converge when dealing with larger magnitudes due to the increased curvature of the loss function. This phenomenon arises because the derivatives of features with larger relative magnitudes tend to be larger, leading to undesirable weight updates. These abnormal updates, in turn, result in a prolonged convergence process, requiring more steps and subsequently escalating the computational burden.</li><li>By scaling the data to fall within the [-1, 1] range, the error function becomes more spherical. Consequently, models trained with transformed or scaled data in the [-1, 1] range exhibit faster convergence, making them quicker and more cost-effective to train. Additionally, the [-1, 1] range provides the highest floating-point precision.</li></ul><p>Another imperative for scaling datasets is the sensitivity of certain machine learning algorithms and techniques to the relative magnitudes of various features. Take, for instance, clustering algorithms like k-means, which employ Euclidean distance as their proximity measure. Failure to scale may lead the algorithm to heavily favour features with comparatively larger magnitudes.</p><p>Furthermore, feature scaling is crucial for the effectiveness of L1 or L2 regularization. The magnitude of weights associated with a feature is influenced by the magnitude of the values of that feature. As a result, different features are impacted differentially by regularization when not properly scaled. Scaling all features to fall within the [-1, 1] range ensures minimal discrepancies in the relative magnitudes of diverse features.</p><h4>3. Some Feature Scaling Models:</h4><ol><li>Normalisation:</li></ol><p>Normalization, commonly referred to as min-max scaling, leverages the minimum and maximum values within a dataset to transform its range. By default, it scales the data between 0 and 1 unless specified otherwise with custom range values. Importantly, this method doesn’t impact outliers but uniformly scales down all values to fit within the specified range.</p><pre># Sample code<br>from sklearn.preprocessing import MinMaxScaler<br><br># Assuming &#39;data&#39; is your dataset<br>scaler = MinMaxScaler()<br>scaled_data = scaler.fit_transform(data)</pre><p>2. Standardisation:</p><p>Standardization, also known as z-score standardization, involves transforming the data to have a mean of 0 and a standard deviation of 1. This method is particularly useful when features in your dataset have different units or follow different scales.</p><pre># Sample code<br>from sklearn.preprocessing import StandardScaler<br><br># Assuming &#39;data&#39; is your dataset<br>scaler = StandardScaler()<br>scaled_data = scaler.fit_transform(data)</pre><h4>4. Practical Examples</h4><p>Let’s pause for a bit and apply these to an example. So, when we apply the above two methods to the NumPy array of :</p><blockquote>np.array([0, 1, 2, 3, 4, 5])</blockquote><p>we obtain the following result:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/768/1*qBOtmuUtIZWyG1Z5mTGCQA.png" /><figcaption>Standardization and normalization</figcaption></figure><p>Note that, in the formulae, <em>x(i) </em>is the current value of the feature, <em>μ</em> is the mean of the feature(x), and <em>σ</em> is the standard deviation of that feature, <em>x(min)</em> is the minimum value of the feature while <em>x(max)</em> is the maximum value of the feature.</p><p>This was more on a theoretical understanding, but let’s also see a practical implementation of this:</p><pre># Normalization<br>import numpy as np<br><br># Sample data<br>data = np.array([2, 5, 10, 20, 50])<br><br># Calculate the minimum and maximum values<br>min_value = data.min()<br>max_value = data.max()<br><br># Normalize the data<br>normalized_data = (data - min_value) / (max_value - min_value)<br><br>print(&quot;Original Data:&quot;, data)<br>print(&quot;Normalized Data:&quot;, normalized_data)</pre><pre>Original Data: [ 2  5 10 20 50]<br>Normalized Data: [0. 0.09 0.27 0.54 1.  ]</pre><p>In this example, we’ve taken a sample dataset and applied Min-Max scaling to normalize it. The values are now within the range [0, 1]. And for standardization, we have:</p><pre>import numpy as np<br><br># Sample data<br>data = np.array([10, 20, 30, 40, 50])<br><br># Calculate the mean and standard deviation<br>mean = data.mean()<br>std_dev = data.std()<br><br># Standardize the data<br>standardized_data = (data - mean) / std_dev<br><br>print(&quot;Original Data:&quot;, data)<br>print(&quot;Standardized Data:&quot;, standardized_data)</pre><pre>Original Data: [10 20 30 40 50]<br>Standardized Data: [-1.26491106 -0.63245553  0. 0.63245553  1.26491106]</pre><p>In this example, we’ve taken a sample dataset and applied z-score standardization to it. The mean is now 0, and the standard deviation is 1, making the data suitable for algorithms that assume a normal distribution.</p><p>There is one more model that I have found a bit interesting but not used pretty commonly in practice robust scaling.</p><h4>5. When to use what?</h4><p>The decision between data normalization and data standardization relies on the nature of your data and analysis requirements.</p><ul><li>You aim to scale features to a specific range, typically [0, 1], without the need for a normal distribution. Examples include image pixel values or scores on a scale from 1 to 100, this is when you opt for normalization.</li><li>Your data should adhere to a standard normal distribution with a mean of 0 and a standard deviation of 1. This is beneficial for algorithms like Principal Component Analysis (PCA) or when dealing with data following a normal distribution, which is why you use standardization.</li></ul><p>To illustrate these distinctions, let’s delve into a practical scenario.</p><p>Imagine you possess a dataset of student exam scores where “Math” scores range from 0 to 100, and “English” scores range from 0 to 800. In this context:</p><p>Math Scores: Utilize data normalization to scale both Math and English scores to a common range, like [0, 1].</p><p>English Scores: Given the significantly different scale of English scores, opt for data standardization to ensure a mean of 0 and a standard deviation of 1, facilitating fair comparison.</p><p>In real-world applications, it’s common to employ diverse preprocessing techniques for different features within the same dataset based on their individual characteristics.</p><h4>6. Challenges</h4><p>Data normalization and standardization, though powerful, present challenges and considerations. Loss of interpretability can arise as the original meaning of data may be obscured during these processes. To address this, it’s advisable to retain the original data for reference.</p><p>Overfitting, a risk associated with complex models, can be exacerbated by preprocessing techniques like standardization. Mitigate this by employing cross-validation and regularization to maintain model robustness.</p><p>Outliers, significantly deviating data points, pose challenges during standardization, affecting mean and standard deviation calculations. Techniques like winsorizing or using robust statistics help address outliers and maintain data integrity.</p><p>While preprocessing aids model performance, it should not replace thoughtful feature engineering. Domain-specific transformations contribute to creating more informative features, impacting model performance significantly.</p><p>Choosing the appropriate preprocessing method is crucial. Experiment with both normalization and standardization, considering the nature of your data and analysis requirements to optimize results.</p><p>In conclusion, feature scaling is a crucial preprocessing step in machine learning, ensuring that each feature contributes equally to model training. Python provides various tools, such as those in the sci-kit-learn library, to implement different scaling techniques based on the characteristics of your dataset and the requirements of your machine-learning algorithm.</p><p>By understanding and utilizing methods like normalization, standardization, and robust scaling, you can enhance the robustness and effectiveness of your machine-learning models. Remember, the choice of scaling method depends on the specific nature of your data and the nuances of the machine learning algorithm you are working with.</p><p><em>Thanks for reading! Do subscribe and share with everyone in your community and don’t forget to like and comment. Follow my Medium Page </em><a href="https://medium.com/u/6521c7f86bab"><em>Rishika Gupta</em></a><em>.</em></p><h3>In Plain English 🚀</h3><p><em>Thank you for being a part of the </em><a href="https://plainenglish.io/"><strong><em>In Plain English</em></strong></a><em> community! Before you go:</em></p><ul><li>Be sure to <strong>clap</strong> and <strong>follow</strong> the writer ️👏<strong>️️</strong></li><li>Follow us: <a href="https://twitter.com/inPlainEngHQ"><strong>X</strong></a><strong> | </strong><a href="https://www.linkedin.com/company/inplainenglish/"><strong>LinkedIn</strong></a><strong> | </strong><a href="https://www.youtube.com/channel/UCtipWUghju290NWcn8jhyAw"><strong>YouTube</strong></a><strong> | </strong><a href="https://discord.gg/in-plain-english-709094664682340443"><strong>Discord</strong></a><strong> | </strong><a href="https://newsletter.plainenglish.io/"><strong>Newsletter</strong></a></li><li>Visit our other platforms: <a href="https://stackademic.com/"><strong>Stackademic</strong></a><strong> | </strong><a href="https://cofeed.app/"><strong>CoFeed</strong></a><strong> | </strong><a href="https://venturemagazine.net/"><strong>Venture</strong></a></li><li>More content at <a href="https://plainenglish.io/"><strong>PlainEnglish.io</strong></a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c16f58f1f017" width="1" height="1" alt=""><hr><p><a href="https://python.plainenglish.io/feature-scaling-c16f58f1f017">Feature Scaling</a> was originally published in <a href="https://python.plainenglish.io/">Python in Plain English</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Web Scraping with Python]]></title>
            <link>https://python.plainenglish.io/web-scraping-with-python-29dbd340b4f5?source=rss-6521c7f86bab------2</link>
            <guid isPermaLink="false">https://medium.com/p/29dbd340b4f5</guid>
            <category><![CDATA[beautifulsoup]]></category>
            <category><![CDATA[data-extraction]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[web-scraping]]></category>
            <dc:creator><![CDATA[Rishika Gupta]]></dc:creator>
            <pubDate>Thu, 12 Oct 2023 20:38:49 GMT</pubDate>
            <atom:updated>2023-10-16T01:01:14.195Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*xuv4XIze-bM1-PK2" /><figcaption>Photo by <a href="https://unsplash.com/@nasa?utm_source=medium&amp;utm_medium=referral">NASA</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Web scraping is a powerful tool for extracting data from websites. Python provides several libraries that make web scraping relatively easy and efficient. In this tutorial, we will cover the basics of web scraping, essential Python libraries, and demonstrate the process with code snippets and examples.</p><h3>Table of Contents</h3><ol><li>Introduction to Web Scraping<br> — What is Web Scraping?<br> — Legality and Ethics</li><li>Basics of HTML and CSS HTML Structure<br> — CSS Selectors<br> — Inspecting a Web Page</li><li>Python Libraries for Web Scraping<br> — Requests<br> — Beautiful Soup<br> — Selenium</li><li>Setting Up the Environment<br> — Installing Required Libraries</li><li>Web Scraping Workflow<br> — Sending HTTP Requests<br> — Parsing HTML Content<br> — Extracting Data</li><li>Handling Different Types of Data<br> — Text Data<br> — Images</li><li>Data Storage<br> — CSV<br> — JSON</li><li>Error Handling and Debugging<br> — Handling HTTP Errors<br> — Debugging Techniques</li><li>Conclusion</li><li>Next steps</li></ol><h3>1. Introduction to Web Scraping</h3><p><strong>What is Web Scraping?</strong><br>Web scraping involves extracting information from websites. It’s essential to do this responsibly, respecting the website’s terms of use and legal boundaries.</p><h4>Legality and Ethics</h4><p>Always ensure you have the right to access and extract data from a website. Adhere to ethical practices and respect the site’s guidelines.</p><h3>2. Basics of HTML and CSS</h3><h4>HTML Structure</h4><p>HTML is the backbone of web pages. It comprises various elements like headings, paragraphs, lists, etc., organized in a structured manner.</p><h4>CSS Selectors</h4><p>CSS selectors help identify HTML elements. They are used to apply styles to these elements and are crucial for targeting specific data during web scraping.</p><h4>Inspecting a Web Page</h4><p>Modern web browsers provide Developer Tools to inspect and analyze a webpage’s structure. These tools help in identifying HTML elements and their associated CSS selectors.</p><h3>3. Python Libraries for Web Scraping</h3><h4>Requests</h4><p>The requests library in Python allows us to send HTTP requests to a URL. It’s a fundamental step in web scraping to fetch the content of a webpage.</p><pre>import requests<br><br>url = &#39;https://example.com&#39;<br><br>response = requests.get(url)<br><br>if response.status_code == 200:<br> print(response.content)<br>else:<br> print(&#39;Failed to retrieve the webpage&#39;)</pre><h4>Beautiful Soup</h4><p>Beautiful Soup is a powerful library for parsing HTML and XML documents. It simplifies navigating and searching the HTML DOM.</p><pre>from bs4 import BeautifulSoup<br><br># Assuming &#39;html_content&#39; contains the HTML of a webpage<br>soup = BeautifulSoup(html_content, &#39;html.parser&#39;)<br><br># Print the page title<br>print(soup.title.text)<br><br># Extract all paragraph texts<br>paragraphs = soup.find_all(&#39;p&#39;)<br>for p in paragraphs:<br> print(p.text)</pre><h4>Selenium</h4><p>Selenium is a web testing framework, but it’s also useful for web scraping dynamic web pages rendered with JavaScript.</p><pre>from selenium import webdriver<br><br>url = &#39;https://example.com&#39;<br>driver = webdriver.Chrome() # or any other WebDriver<br>driver.get(url)<br><br># Access elements through CSS selectors<br>element = driver.find_element_by_css_selector(&#39;.example-class&#39;)<br>print(element.text)<br><br>driver.quit()</pre><h3>4. Setting Up the Environment</h3><h4>Installing Required Libraries</h4><p>Install the necessary libraries using pip:</p><pre>pip install requests beautifulsoup4 selenium</pre><h3>5. Web Scraping Workflow</h3><h4>Sending HTTP Requests</h4><p>Sending HTTP requests is the first step in web scraping. Use the requests library to initiate a request and retrieve the HTML content of the webpage.</p><pre>import requests<br><br>url = &#39;https://example.com&#39;<br>response = requests.get(url)<br><br>if response.status_code == 200:<br> # Print the content of the webpage<br> print(response.content)<br>else:<br> print(&#39;Failed to retrieve the webpage&#39;)</pre><h4>Parsing HTML Content</h4><p>After obtaining the HTML content, use Beautiful Soup to parse the HTML and navigate the Document Object Model (DOM).</p><pre>from bs4 import BeautifulSoup<br><br># Assuming &#39;html_content&#39; contains the HTML of a webpage<br>soup = BeautifulSoup(html_content, &#39;html.parser&#39;)  <br><br># Print the page title<br>print(soup.title.text)</pre><h4>Extracting Data</h4><p>Once the HTML is parsed, use various methods provided by Beautiful Soup to extract the data you need from specific HTML elements.</p><pre># Extract all paragraph texts<br>paragraphs = soup.find_all(‘p’)<br>for p in paragraphs:<br> print(p.text)</pre><h3>6. Handling Different Types of Data</h3><h4>Text Data</h4><p>Extracting text data from HTML elements is a common task. Use Beautiful Soup to target and retrieve text.</p><pre># Assuming ‘element’ is an HTML element<br>text = element.text<br>print(text)</pre><h4>Images</h4><p>To download images, extract the image URLs from HTML elements and use the requests library to download and store them locally.</p><pre># Assuming ‘img_element’ is an HTML img element<br>image_url = img_element[&#39;src&#39;]<br><br># Download the image<br>img_data = requests.get(image_url).content<br>with open(&#39;image.jpg&#39;, &#39;wb&#39;) as f:<br> f.write(img_data)</pre><h3>7. Data Storage</h3><h4>CSV</h4><p>Storing data in CSV format is simple and effective.</p><pre>import csv<br><br>data = [{&#39;name&#39;: &#39;Alice&#39;, &#39;age&#39;: 30}, {&#39;name&#39;: &#39;Bob&#39;, &#39;age&#39;: 35}]<br><br># Write to CSV<br>with open(&#39;data.csv&#39;, &#39;w&#39;, newline=&#39;&#39;) as csvfile:<br> fieldnames = [&#39;name&#39;, &#39;age&#39;]<br> writer = csv.DictWriter(csvfile, fieldnames=fieldnames)<br> <br> writer.writeheader()<br> for item in data:<br> writer.writerow(item)</pre><h4>JSON</h4><p>Storing data in JSON format is also straightforward.</p><pre>import json<br><br>data = [{&#39;name&#39;: &#39;Alice&#39;, &#39;age&#39;: 30}, {&#39;name&#39;: &#39;Bob&#39;, &#39;age&#39;: 35}]<br><br># Write to JSON<br>with open(&#39;data.json&#39;, &#39;w&#39;) as json_file:<br> json.dump(data, json_file)</pre><h3>8. Error Handling and Debugging</h3><h4>Handling HTTP Errors</h4><p>Implement error handling mechanisms to handle HTTP errors gracefully.</p><pre>import requests<br><br>url = &#39;https://example.com&#39;<br>response = requests.get(url)<br><br>if response.status_code == 200:<br> print(response.content)<br>else:<br> print(&#39;Failed to retrieve the webpage. Status code:&#39;, response.status_code)</pre><h4>Debugging Techniques</h4><p>Use debugging techniques to identify and resolve issues.</p><pre># Print statements for debugging<br>print(‘Debug point 1’)<br>print(variable)<br></pre><h3>9. Conclusion and Best Practices</h3><p>In this tutorial, we’ve covered the fundamentals of web scraping using Python, including understanding HTML and CSS, using essential Python libraries, and demonstrating the web scraping workflow. Before we conclude, let’s discuss some best practices to ensure effective and ethical web scraping:</p><h4>Respect the Website’s Terms of Use</h4><p>Always review and comply with the website’s terms of use and policies. Ensure that your web scraping activities align with the website’s guidelines to maintain ethical behavior.</p><h4>Use a User-Agent Header</h4><p>Set a user-agent header in your HTTP requests to identify your scraper as a legitimate user. Some websites may block requests without a user-agent header or with suspicious user-agents.</p><pre>headers = {&quot;User-Agent&quot;: &quot;Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.3&quot;}<br>response = requests.get(url, headers=headers)</pre><h4>Respect the rules</h4><p>Adhere to the rules specified on the website, which can provide instructions to web crawlers regarding what pages to access and scrape. Always follow these guidelines to maintain good scraping behavior.</p><h4>Implement Rate Limiting</h4><p>Implement rate limiting to control the frequency of your requests and avoid overwhelming the target server. Excessive requests in a short period can lead to IP bans or temporary blocks.</p><pre>import time<br><br># Delay between requests (in seconds)<br>delay = 2<br><br># Loop through URLs and scrape with delay<br>for url in urls:<br> response = requests.get(url)<br> # Process response<br> time.sleep(delay) # Delay before the next request</pre><h4>Handle Errors Gracefully</h4><p>Implement proper error handling to handle potential issues such as HTTP errors, timeouts, or network problems. This ensures your scraper can recover from errors and continue its operation smoothly.</p><pre>try:<br> response = requests.get(url)<br> response.raise_for_status()<br>except requests.exceptions.HTTPError as errh:<br> print(“HTTP Error:”, errh)<br>except requests.exceptions.ConnectionError as errc:<br> print(“Error Connecting:”, errc)<br>except requests.exceptions.Timeout as errt:<br> print(“Timeout Error:”, errt)<br>except requests.exceptions.RequestException as err:<br> print(“Something went wrong:”, err)</pre><h3>10. Next Steps</h3><p>Now that you have a solid understanding of web scraping with Python, you can explore more advanced topics, experiment with different websites, and enhance your skills. Consider diving deeper into handling JavaScript-rendered pages, authentication and sessions, using proxies, and optimizing your scraping techniques for specific use cases.</p><p>Keep in mind that responsible and ethical web scraping is essential to maintain a positive and respectful presence on the internet. Always abide by the terms of use and respect the guidelines set by the websites you scrape.</p><p>Happy web scraping!</p><p><em>Thanks for reading! Do subscribe and share with everyone in your community and don’t forget to like and comment. Follow my Medium Page </em><a href="https://medium.com/u/6521c7f86bab"><em>Rishika Gupta</em></a><em>.</em></p><h3>In Plain English</h3><p><em>Thank you for being a part of our community! Before you go:</em></p><ul><li><em>Be sure to </em><strong><em>clap</em></strong><em> and </em><strong><em>follow</em></strong><em> the writer! 👏</em></li><li><em>You can find even more content at </em><a href="https://plainenglish.io/"><strong><em>PlainEnglish.io</em></strong></a><strong><em> 🚀</em></strong></li><li><em>Sign up for our </em><a href="http://newsletter.plainenglish.io/"><strong><em>free weekly newsletter</em></strong></a><em>. 🗞️</em></li><li><em>Follow us on </em><a href="https://twitter.com/inPlainEngHQ"><strong><em>Twitter</em></strong></a><strong><em>(X</em></strong>), <a href="https://www.linkedin.com/company/inplainenglish/"><strong><em>LinkedIn</em></strong></a>, <a href="https://www.youtube.com/channel/UCtipWUghju290NWcn8jhyAw"><strong><em>YouTube</em></strong></a>, and <a href="https://discord.gg/in-plain-english-709094664682340443"><strong><em>Discord</em></strong></a><strong><em>.</em></strong></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=29dbd340b4f5" width="1" height="1" alt=""><hr><p><a href="https://python.plainenglish.io/web-scraping-with-python-29dbd340b4f5">Web Scraping with Python</a> was originally published in <a href="https://python.plainenglish.io/">Python in Plain English</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Reinforcement Learning: What’s all this?]]></title>
            <link>https://ai.plainenglish.io/reinforcement-learning-whats-all-this-f83d51017416?source=rss-6521c7f86bab------2</link>
            <guid isPermaLink="false">https://medium.com/p/f83d51017416</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[reinforcement-learning]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[algorithms]]></category>
            <dc:creator><![CDATA[Rishika Gupta]]></dc:creator>
            <pubDate>Thu, 12 Oct 2023 20:02:24 GMT</pubDate>
            <atom:updated>2023-10-13T01:06:02.642Z</atom:updated>
            <content:encoded><![CDATA[<h3>Understanding Reinforcement Learning: A Comprehensive Guide</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*VpmPez2jo_iJ7QUc" /><figcaption>Photo by <a href="https://unsplash.com/@santesson89?utm_source=medium&amp;utm_medium=referral">Andrea De Santis</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>In the vast realm of artificial intelligence, Reinforcement Learning (RL) stands as a prominent pillar, offering a unique approach to machine learning. It differs significantly from both supervised and unsupervised learning, presenting a distinctive learning paradigm. In this article, we’ll delve into the fundamental concepts of reinforcement learning, elucidate its differences in comparison to supervised and unsupervised learning, and provide insightful examples to aid in understanding.</p><h4>Table of Contents</h4><ol><li>Introduction to Reinforcement Learning</li><li>Reinforcement Learning vs. Supervised Learning</li><li>Reinforcement Learning vs. Unsupervised Learning</li><li>Understanding the Reinforcement Learning Process</li><li>Practical Examples</li><li>Conclusion</li></ol><h4>Brief Introduction to Reinforcement Learning:</h4><p>Reinforcement Learning is a type of machine learning where an agent learns how to behave in an environment by performing actions and receiving rewards or penalties. The goal is to maximize the cumulative reward the agent receives over time by learning optimal strategies or policies.</p><p>In RL, the agent interacts with the environment, observes its state, takes actions, and receives feedback in the form of rewards or punishments based on its actions. The agent learns by trial and error, aiming to find the best possible actions to achieve a specific objective.</p><h4>Reinforcement Learning vs. Supervised Learning:</h4><p>In supervised learning, the model is trained on a labeled dataset, where each input has a corresponding output. The model learns to map inputs to outputs based on the provided labels. On the other hand, reinforcement learning involves learning through interaction with the environment, where the model learns optimal behavior through experimentation and feedback.</p><p>The key difference is that in supervised learning, the model is provided explicit instructions (labels) on how to map inputs to outputs, whereas in reinforcement learning, the model learns optimal actions by trial and error, guided by rewards.</p><h4>Reinforcement Learning vs. Unsupervised Learning:</h4><p>Unsupervised learning involves discovering patterns and structures in data without explicit supervision or labelled outputs. It often includes tasks like clustering and dimensionality reduction. In contrast, reinforcement learning focuses on learning optimal actions to maximize a reward signal by interacting with an environment.</p><p>In unsupervised learning, the model learns from the inherent structure of the data, while reinforcement learning learns from the dynamic interaction with the environment and the resulting feedback.</p><h4>Understanding the Reinforcement Learning Process:</h4><p>The reinforcement learning process typically involves the following components:</p><ul><li><strong>Agent</strong>: The learner or decision-maker that interacts with the environment.</li><li><strong>Environment</strong>: The external system with which the agent interacts, presenting different states and responses to the agent’s actions.</li><li><strong>State</strong>: A representation of the current situation or configuration of the environment that the agent observes.</li><li><strong>Action</strong>: The decision or move made by the agent based on the observed state.</li><li><strong>Reward</strong>: The feedback received by the agent after taking an action in a particular state. It indicates the immediate benefit or penalty associated with the action.</li><li><strong>Policy</strong>: The strategy or function that defines the agent’s behavior, specifying the mapping between states and actions.</li></ul><p>The objective of reinforcement learning is to find the optimal policy that maximizes the cumulative reward over time.</p><h4>Practical Examples:</h4><p><strong>Playing Chess: </strong>Consider an RL agent learning to play chess. The state represents the current board position, actions correspond to legal moves, and the reward could be a positive value for capturing an opponent’s piece or a negative value for losing a piece. The agent learns a policy that guides its moves to win the game.</p><p><strong>Robot Navigation: </strong>In robot navigation, the environment is the physical space, states are different positions in the space, actions are movements (e.g., forward, backward, left, right), and rewards could be positive for reaching the destination or negative for collisions. The agent learns a policy to navigate efficiently.</p><h3>Conclusion:</h3><p>Reinforcement Learning offers a unique learning paradigm by allowing an agent to learn optimal behavior through interaction with an environment and feedback in the form of rewards or penalties. Contrasting it with supervised and unsupervised learning helps highlight its distinctive characteristics. By understanding the reinforcement learning process and exploring practical examples, we can appreciate its vast potential and versatility across various domains.</p><p>Embark on your journey into the world of reinforcement learning, explore its applications, and uncover the endless possibilities it offers in the realm of artificial intelligence. Happy learning!</p><p><em>Thanks for reading! Do subscribe and share with everyone in your community and don’t forget to like and comment. Follow my Medium Page </em><a href="https://medium.com/u/6521c7f86bab"><em>Rishika Gupta</em></a><em>.</em></p><h3>In Plain English</h3><p><em>Thank you for being a part of our community! Before you go:</em></p><ul><li><em>Be sure to </em><strong><em>clap</em></strong><em> and </em><strong><em>follow</em></strong><em> the writer! 👏</em></li><li><em>You can find even more content at </em><a href="https://plainenglish.io/"><strong><em>PlainEnglish.io</em></strong></a><strong><em> 🚀</em></strong></li><li><em>Sign up for our </em><a href="http://newsletter.plainenglish.io/"><strong><em>free weekly newsletter</em></strong></a><em>. 🗞️</em></li><li><em>Follow us on </em><a href="https://twitter.com/inPlainEngHQ"><strong><em>Twitter</em></strong></a><strong><em>(X</em></strong>), <a href="https://www.linkedin.com/company/inplainenglish/"><strong><em>LinkedIn</em></strong></a>, <a href="https://www.youtube.com/channel/UCtipWUghju290NWcn8jhyAw"><strong><em>YouTube</em></strong></a>, and <a href="https://discord.gg/in-plain-english-709094664682340443"><strong><em>Discord</em></strong></a><strong><em>.</em></strong></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f83d51017416" width="1" height="1" alt=""><hr><p><a href="https://ai.plainenglish.io/reinforcement-learning-whats-all-this-f83d51017416">Reinforcement Learning: What’s all this?</a> was originally published in <a href="https://ai.plainenglish.io/">Artificial Intelligence in Plain English</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[An introduction to Linux]]></title>
            <link>https://medium.datadriveninvestor.com/an-introduction-to-linux-81f45256efc7?source=rss-6521c7f86bab------2</link>
            <guid isPermaLink="false">https://medium.com/p/81f45256efc7</guid>
            <category><![CDATA[linux]]></category>
            <category><![CDATA[unix]]></category>
            <category><![CDATA[architecture]]></category>
            <category><![CDATA[history]]></category>
            <category><![CDATA[computing]]></category>
            <dc:creator><![CDATA[Rishika Gupta]]></dc:creator>
            <pubDate>Wed, 20 Sep 2023 18:46:51 GMT</pubDate>
            <atom:updated>2023-10-05T03:09:31.998Z</atom:updated>
            <content:encoded><![CDATA[<h3>An Introduction to Linux</h3><p>In the vast realm of computer technology, operating systems (OS) are the fundamental bridges between hardware and software, enabling seamless interactions and efficient resource management. One prominent family of operating systems is Unix, which has birthed various popular Unix-based OS such as HP-UX, IBM AIX, Apple macOS, and Oracle Solaris. Unix’s rich history, dating back to the 1960s, and its evolution into diverse branches like Linux, contribute significantly to the computing landscape we know today.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nCx4WtdvdwlzKnww-RhhLg.jpeg" /><figcaption>An introduction to Linux</figcaption></figure><h4>Unraveling the History of Unix</h4><p>Originally conceived in the 1960s at AT&amp;T Labs, Unix was tailored for the PDP-7 computer, a specific hardware system. However, its true breakthrough came in the 1970s when it was rewritten in C, making it portable across multiple hardware architectures. Subsequently, UC Berkeley’s development of the Berkeley Software Distribution (BSD) added further capabilities to Unix. Apple’s macOS traces its lineage back to BSD, showcasing the lasting influence of Unix.</p><h4>The Linux Revolution</h4><p>Linux, a family of Unix-like operating systems, gained prominence as a free and open-source alternative to Unix. Linus Torvalds, in 1991, spearheaded the development of the Linux kernel, a critical component that allows communication between the operating system and machine hardware. This project grew steadily and merged with GNU tools in 1992, creating a powerful, collaborative, and adaptable OS. The Linux penguin, known as TUX, emerged as its endearing mascot, highlighting the friendly and innovative spirit of the Linux community.</p><h4>Linux in the Modern World</h4><p>In contemporary society, Linux operates in diverse domains, showcasing its versatility. From smartphones (with Android being Linux-based) to supercomputers (leveraging clustered Linux-powered servers), from data centres and cloud services to personal computers, Linux has become an integral part of the technology landscape. Its various distributions, known as “distros,” cater to specific tasks and audiences, offering a range of system utilities, graphical interfaces, shell commands, and distinct support models.</p><h3>Exploring Linux Distributions: A Glimpse into Debian, Ubuntu, Red Hat, Fedora, and Arch Linux</h3><h4>Debian: A Pioneer in Open Source</h4><p>Debian, one of the earliest Linux distributions, emerged in 1993, with a stable release in 1996. It gained prominence for its stability, reliability, and extensive community support. Debian’s commitment to open-source principles and its compatibility across diverse computer architectures make it a prominent choice among users.</p><h4>Ubuntu: Building on Debian’s Legacy</h4><p>Rooted in Debian, Ubuntu made its official debut in 2004 and was a significant milestone in Linux’s popularity. Developed and managed by Canonical, Ubuntu offers three versions: Desktop, Server, and Core, catering to a broad spectrum of users. Leveraging Debian’s tools and infrastructure, Ubuntu provides a user-friendly experience while upholding open-source principles.</p><h4>Red Hat Linux: A Focus on Enterprise</h4><p>Red Hat Linux, another stalwart in the Linux landscape, forms the core of the popular Red Hat Enterprise Linux (RHEL). Renowned for its stability, reliability, and robust enterprise focus, RHEL is managed by Red Hat, an IBM subsidiary. Its targeted approach towards enterprise customers underscores its position as a dominant player in the corporate world.</p><h4>Fedora: Balancing Reliability and Innovation</h4><p>Fedora, a stable and secure OS, is recognized for its robustness and unique firewall and security features. Supported by a thriving community and sponsored by Red Hat, Fedora strikes a balance between stability and cutting-edge software, making it a preferred choice for various users.</p><h4>Arch Linux: The DIY Linux</h4><p>Arch Linux appeals to tinkerers and enthusiasts, offering a DIY approach for customizing every aspect of the system. Its flexibility and leading-edge software attract users with a deep understanding of Linux and system tools, making it a go-to choice for those seeking tailor-made solutions.</p><h3>Unveiling the Inner Workings of Linux: Exploring Its Architecture and Functionality</h3><p>In the intricate world of operating systems, Linux stands out as a powerful and versatile choice, powering a myriad of devices we interact with daily. Beneath its user-friendly exterior lies a robust architecture, orchestrating seamless experiences and efficient resource management. To truly grasp its capabilities, we need to delve into the layers and functionalities that define the Linux architecture and explore how it interacts with hardware to deliver a streamlined computing environment.</p><h3>Linux Architecture: A Multilayered Approach</h3><p>Linux’s architecture is structured into five layers, each playing a pivotal role in the system’s operation and user interaction.</p><h4>1. UI (User Interface) Layer — The Outermost Facade</h4><p>The User Interface (UI) layer, residing at the top, serves as the interface between users and the system. It allows users to interact with the system using peripherals such as a keyboard or mouse. In desktop versions, this layer encompasses a Graphical User Interface (GUI), providing an intuitive platform for users to navigate and operate the system. Tasks like sending emails through a web browser or listening to music using a player are made possible through the UI.</p><h4>2. Application Layer — Enabling Functionality</h4><p>Directly below the UI layer is the Application Layer, the realm where specific tasks and functionalities come to life. Here, a vast array of software resides, ranging from essential system tools like compilers and programming languages to user applications such as browsers, text editors, and games. This layer is the powerhouse of functionality, enabling users to perform diverse tasks efficiently.</p><h4>3. Operating System Layer — The Core of Functionality</h4><p>Sitting at the heart of the Linux architecture is the Operating System (OS) layer. This crucial layer controls vital processes and programs, essential for the system’s health, stability, and functionality. Among its myriad functions are software assignment to users, file management, error detection, and failure prevention. The OS layer acts as the orchestrator, ensuring a smooth and secure user experience.</p><h4>4. Kernel — The Vital Core</h4><p>At the core of the Linux architecture lies the Kernel, an indispensable component responsible for managing essential system functions. Serving as the lowest level of software, the Kernel holds complete control over the system and remains in computer memory while the system is operational. It acts as a bridge between applications and hardware, enabling seamless communication through system calls. Memory management, process management, device driver management for hardware support, and ensuring system security are among its primary functions.</p><h4>5. Hardware Layer — The Foundation</h4><p>The foundation of the Linux architecture is the Hardware Layer, comprising the physical and electronic devices that constitute the computer. This includes the Central Processing Unit (CPU), which executes most calculations, Random Access Memory (RAM) for temporary storage, storage devices for data persistence, the computer screen, and various USB devices such as keyboards, mice, and USB drives.</p><h4>The Kernel: A Closer Look</h4><p>Within the Linux architecture, the Kernel stands as a pivotal entity, facilitating the management of the system and ensuring efficient communication between software and hardware. Serving as a bridge between applications and hardware, the Kernel plays a vital role in crucial tasks such as memory management, process management, device driver management for hardware support, and system security. This core component is fundamental to the Linux experience, ensuring optimal performance and stability.</p><h3>Conclusion</h3><p>Operating systems have come a long way since the inception of Unix in the 1960s. The evolution of Unix into diverse branches like Linux has revolutionized the digital landscape, influencing everything from smartphones to supercomputers. As we delve deeper into the world of Linux and its distributions, we appreciate the vast array of choices available to users, catering to their specific needs and preferences. Whether it’s the stability of Debian, the user-friendliness of Ubuntu, the enterprise focus of Red Hat Linux, the innovative approach of Fedora, or the customizability of Arch Linux, Linux distributions continue to shape the digital world, offering a diverse and exciting future for technology enthusiasts and professionals alike.</p><p>Furthermore, understanding the architecture of Linux is akin to unravelling the intricate mechanisms of a finely tuned machine. The layers within, each with its own unique function, blend seamlessly to deliver a robust and user-friendly experience. From the outermost User Interface (UI) layer to the indispensable Kernel, Linux’s architecture ensures efficient resource management, secure operations, and a platform for unparalleled user interactions. Armed with this knowledge, users and developers can harness the full potential of Linux, creating powerful applications and optimizing performance across a vast array of devices.</p><p><em>Thanks for reading! Do subscribe and share with everyone in your community and don’t forget to like and comment. Follow my Medium Page </em><a href="https://medium.com/u/6521c7f86bab"><em>Rishika Gupta</em></a><em>.</em></p><p>Subscribe to DDIntel <a href="https://ddintel.datadriveninvestor.com/">Here</a>.</p><p><a href="https://ddintel.datadriveninvestor.com/"><em>DDIntel</em> </a>captures the more notable pieces from our <a href="https://www.datadriveninvestor.com/"><em>main site</em></a> and our popular <a href="https://medium.datadriveninvestor.com/"><em>DDI Medium publication</em></a>. Check us out for more insightful work from our community.</p><p><a href="https://www.aitoolverse.com/register">Register </a>on AItoolverse (alpha) to get 50 DDINs</p><p>Join our network here: <a href="https://datadriveninvestor.com/collaborate">https://datadriveninvestor.com/collaborate</a></p><p>DDI Official Telegram Channel: <a href="https://t.me/+tafUp6ecEys4YjQ1">https://t.me/+tafUp6ecEys4YjQ1</a></p><p>Follow us on <a href="https://www.linkedin.com/company/data-driven-investor"><em>LinkedIn</em></a>, <a href="https://twitter.com/@DDInvestorHQ"><em>Twitter</em></a>, <a href="https://www.youtube.com/c/datadriveninvestor"><em>YouTube</em></a>, and <a href="https://www.facebook.com/datadriveninvestor"><em>Facebook</em></a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=81f45256efc7" width="1" height="1" alt=""><hr><p><a href="https://medium.datadriveninvestor.com/an-introduction-to-linux-81f45256efc7">An introduction to Linux</a> was originally published in <a href="https://medium.datadriveninvestor.com/">DataDrivenInvestor</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Demystifying Artificial Intelligence and Machine Learning]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://ai.plainenglish.io/demystifying-artificial-intelligence-and-machine-learning-3715430d1be6?source=rss-6521c7f86bab------2"><img src="https://cdn-images-1.medium.com/max/2600/0*yB_SHs7v4EVn7ok1" width="2765"></a></p><p class="medium-feed-snippet">Delve into the nuances of AI vs. ML that set them apart and grasp its impact on some real-world applications</p><p class="medium-feed-link"><a href="https://ai.plainenglish.io/demystifying-artificial-intelligence-and-machine-learning-3715430d1be6?source=rss-6521c7f86bab------2">Continue reading on Artificial Intelligence in Plain English »</a></p></div>]]></description>
            <link>https://ai.plainenglish.io/demystifying-artificial-intelligence-and-machine-learning-3715430d1be6?source=rss-6521c7f86bab------2</link>
            <guid isPermaLink="false">https://medium.com/p/3715430d1be6</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[real-life]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[technical-writing]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Rishika Gupta]]></dc:creator>
            <pubDate>Fri, 15 Sep 2023 15:00:51 GMT</pubDate>
            <atom:updated>2023-09-19T08:11:26.972Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Navigating the BDMA Erasmus Mundus Master’s Programme]]></title>
            <link>https://rishikagupta1999.medium.com/navigating-the-bdma-erasmus-mundus-masters-programme-cd2edb306d9c?source=rss-6521c7f86bab------2</link>
            <guid isPermaLink="false">https://medium.com/p/cd2edb306d9c</guid>
            <category><![CDATA[analytics]]></category>
            <category><![CDATA[erasmus]]></category>
            <category><![CDATA[masters-degree]]></category>
            <category><![CDATA[big-data]]></category>
            <category><![CDATA[erasmus-plus]]></category>
            <dc:creator><![CDATA[Rishika Gupta]]></dc:creator>
            <pubDate>Sat, 02 Sep 2023 23:46:11 GMT</pubDate>
            <atom:updated>2023-10-07T07:46:04.191Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/351/1*whBpTynb2shxVglFAopwhw.png" /><figcaption>BDMA: Big Data Management and Analytics</figcaption></figure><p>A Gateway to Big Data Excellence - Welcome to the BDMA (Big Data Management and Analytics) Erasmus Mundus Program! As the Erasmus Mundus Association Programme Representative of BDMA, I have been inundated with inquiries from aspiring students seeking to embark on this extraordinary educational journey. In response to the overwhelming interest and curiosity, I have taken it upon myself to create this comprehensive article, aimed at shedding light on every aspect of the BDMA program — from its inception to the application process, tips for success, and what to expect during your BDMA experience.</p><h3>Programme Overview</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*S38y1upfpSCFHoJ-" /><figcaption>Photo by <a href="https://unsplash.com/@helloimnik?utm_source=medium&amp;utm_medium=referral">Nik</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h4>Pioneering Excellence in Big Data</h4><p>BDMA, which commenced in the year 2012 (previously as IT4BI), stands as a testament to the relentless pursuit of excellence in the realm of Big Data and Data Science. As I am currently enrolled in the XYZ edition, I can confidently attest to the program’s outstanding curriculum and unique characteristics.</p><p>If you are looking for the official source of information, the BDMA website, <a href="https://bdma.ulb.ac.be/">https://bdma.ulb.ac.be/</a>, serves as your portal into the world of BDMA. The website is a treasure trove of information, offering a comprehensive look at the program’s structure, faculty, admission details, and much more. For a detailed breakdown of the course content and what you can expect from this transformative journey, visit <a href="https://bdma.ulb.ac.be/home/master-programme/course-content/">https://bdma.ulb.ac.be/home/master-programme/course-content/</a>.</p><h3>Application Process</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*vRpWKj-MVa1Iks5k" /><figcaption>Photo by <a href="https://unsplash.com/@mashaoliynyk?utm_source=medium&amp;utm_medium=referral">Mariya Oliynyk</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h4>Your Ticket to BDMA</h4><p>One of the crucial aspects of realizing your BDMA dream is successfully navigating the application process. While it may seem straightforward, careful attention to detail is paramount.</p><p>Applications can be submitted via the <a href="https://cs.ulb.ac.be/emundus/">BDMA application portal</a>, whether you are applying for a scholarship (before the December deadline) or choosing the self-funding option (before the March deadline). The flexibility in funding options ensures that BDMA caters to a diverse range of students, committed to the study of Big Data and Data Science.</p><p>BDMA is not exclusive to a single profile; it embraces a diverse array of backgrounds and talents. To understand if your profile aligns with the program’s expectations, visit <a href="https://bdma.ulb.ac.be/home/students/bdma-expected-profiles-pre-requisites/">https://bdma.ulb.ac.be/home/students/bdma-expected-profiles-pre-requisites/</a>.</p><p>An intriguing element of the BDMA application process is the requirement to submit specific documents by postal mail (s-mail) in addition to the online uploads. Ensure that you do not overlook this vital step, as it distinguishes BDMA from other master’s programs.</p><p>The standard application requirements encompass a series of documents, including a recent photo, passport, CV, transcripts, and your bachelor’s degree certificate (with a minimum of 180 ECTS credits). If you hold a previous master’s degree, it must also be included in your application. Furthermore, IELTS/TOEFL scores, work experience certificates, a well-crafted cover letter, a self-evaluation sheet provided by the program, and letters of recommendation are essential components of the application package.</p><h3>Tips for a Successful Application</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*gsPdpw-NZdukIglY" /><figcaption>Photo by <a href="https://unsplash.com/@thoughtcatalog?utm_source=medium&amp;utm_medium=referral">Thought Catalog</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h4>Navigating the Path to BDMA</h4><p>To ensure a smooth journey through the application process, consider these tips:</p><ol><li><strong>Consult the Application Manual</strong>: The <a href="https://cs.ulb.ac.be/emundus/images/stories/provided_documents/bdma_application_manual.pdf">BDMA application manual</a> is a comprehensive resource that covers every detail you need to know for a successful application. Treat it as your roadmap.</li><li><strong>Deadline Vigilance</strong>: Keep a watchful eye on the application deadline. Ensure that all necessary documents and application details are submitted punctually. This includes the crucial letters of recommendation, which should be uploaded by your recommender.</li><li><strong>Europass Format</strong>: It is highly recommended to use the Europass format for both your Cover Letter and CV. This widely recognized format ensures professionalism and clarity.</li><li><strong>Crafting Your Cover Letter</strong>: When crafting your cover letter, remember to:</li></ol><ul><li>Be concise, keeping it within one to two pages.</li><li>Articulate how this master’s program will contribute to your personal and professional growth.</li><li>Express your genuine enthusiasm for BDMA and highlight its unique advantages, such as its diverse student body and cutting-edge research facilities.</li><li>Outline your short-term and long-term goals following your BDMA journey.</li><li>Present a compelling narrative, avoiding clichéd structures.</li><li>Showcase your skills, acquired through projects and competitions, and explain how they will enhance your experience in BDMA.</li><li>Don’t hesitate to include additional information that you believe would be relevant to the BDMA selection committee.</li></ul><h3>Future Insights</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*PUAvGduhGWEOktCs" /><figcaption>Photo by <a href="https://unsplash.com/@neom?utm_source=medium&amp;utm_medium=referral">NEOM</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h4>A Glimpse into BDMA Life</h4><p>For those eager to gain insights into life as a BDMA student, I’m excited to announce that I will be publishing an upcoming article. This article will feature detailed experiences shared by current and past BDMA students, offering you invaluable insights to help you make an informed decision about joining this remarkable program.</p><h3>Conclusion</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*7SbIKRBI_9TyAxJ8" /><figcaption>Photo by <a href="https://unsplash.com/@hannynaibaho?utm_source=medium&amp;utm_medium=referral">Hanny Naibaho</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><h4>Connect with BDMA</h4><p>If you have any further questions or require additional information about BDMA, please do not hesitate to reach out to me via LinkedIn (<a href="https://www.linkedin.com/in/rishika-g/">https://www.linkedin.com/in/rishika-g/</a>) or email (<a href="/service/mailto:bdma@em-a.eu">bdma@em-a.eu</a>). Your inquiries are welcome and valuable.</p><p>Thank you for taking the time to read this comprehensive guide. I encourage you to subscribe to stay updated, share this article within your community, and actively engage with it through likes and comments. For more insightful content, be sure to follow my Medium Page — <a href="https://medium.com/u/6521c7f86bab"><em>Rishika Gupta</em></a><em>.</em></p><p>Wishing you the best of luck with your BDMA application journey! May it be the first step towards an enriching and transformative experience in the world of Big Data and Data Science.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cd2edb306d9c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Unlocking Global Horizons: The Erasmus Mundus Scholarship Experience]]></title>
            <link>https://rishikagupta1999.medium.com/unlocking-global-horizons-the-erasmus-mundus-scholarship-experience-7f0d78a48aa1?source=rss-6521c7f86bab------2</link>
            <guid isPermaLink="false">https://medium.com/p/7f0d78a48aa1</guid>
            <category><![CDATA[higher-education]]></category>
            <category><![CDATA[scholarships-for-students]]></category>
            <category><![CDATA[scholarship]]></category>
            <category><![CDATA[education]]></category>
            <category><![CDATA[erasmus]]></category>
            <dc:creator><![CDATA[Rishika Gupta]]></dc:creator>
            <pubDate>Thu, 31 Aug 2023 08:34:07 GMT</pubDate>
            <atom:updated>2023-08-31T08:34:07.380Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*UiDuhsTdQIEr9iNB" /><figcaption>Photo by <a href="https://unsplash.com/@christianlue?utm_source=medium&amp;utm_medium=referral">Christian Lue</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>In a world characterized by increasing globalization, the importance of international education cannot be overstated. Studying abroad not only exposes students to diverse cultures and perspectives but also enhances their academic and personal growth. The Erasmus Mundus Scholarship, a prestigious and sought-after program, offers an exceptional opportunity for students to pursue their higher education dreams across multiple countries. In this article, we delve into the essence of the Erasmus Mundus Scholarship, exploring its benefits, application process, and the life-changing experiences it provides.</p><h4>The Erasmus Mundus Scholarship: A Brief Overview</h4><p>The Erasmus Mundus Scholarship is an initiative by the European Commission to promote international education and cooperation through joint master’s and doctoral programs offered by consortia of European universities. Named after the renowned philosopher Desiderius Erasmus and his travel-intensive life, the program aims to cultivate global citizens who can navigate cultural complexities and contribute to a more interconnected world.</p><h4>Benefits of the Erasmus Mundus Scholarship</h4><ol><li><strong>Academic Excellence</strong>: Erasmus Mundus programs are designed to provide top-tier education through collaboration between multiple institutions. Scholars have access to esteemed faculty members, cutting-edge research, and comprehensive curricula that span various facets of their chosen field.</li><li><strong>Cultural Diversity:</strong> One of the most enriching aspects of the scholarship is the chance to immerse oneself in diverse cultures. Scholars interact with students from all corners of the globe, fostering cross-cultural friendships and broadening their worldview.</li><li><strong>Multiple Destinations:</strong> Erasmus Mundus programs often involve studying in multiple countries, allowing scholars to experience different educational systems and research environments. This exposure adds depth to their academic journey and enriches their personal experiences.</li><li><strong>Language Proficiency:</strong> Living in different countries offers an excellent opportunity to learn new languages or enhance existing language skills. This linguistic diversity not only aids in daily life but also contributes to a more versatile skill set.</li><li><strong>Professional Network:</strong> Scholars are exposed to a wide range of international academic and professional networks. These connections prove invaluable when seeking career opportunities across the globe.</li><li><strong>Enhanced Employability:</strong> Graduates of Erasmus Mundus programs are highly sought after by employers due to their adaptability, intercultural competence, and rigorous academic training.</li></ol><h4>The Application Process</h4><p>The application process for the Erasmus Mundus Scholarship is competitive but manageable with careful planning and preparation.</p><ol><li><strong>Program Research:</strong> Begin by researching the available Erasmus Mundus programs. Each program has its own unique focus, structure, and admission requirements. Select a program that aligns with your academic and career aspirations. The catalogue of programs is available at <a href="https://www.eacea.ec.europa.eu/scholarships/erasmus-mundus-catalogue_en">https://www.eacea.ec.europa.eu/scholarships/erasmus-mundus-catalogue_en</a>.</li><li><strong>Application Compilation:</strong> Prepare a compelling application package that typically includes academic transcripts, recommendation letters, a statement of purpose, and sometimes language proficiency test scores (such as IELTS or TOEFL).</li><li><strong>Online Application:</strong> Apply directly to the chosen Erasmus Mundus program through their online application portal. Be attentive to deadlines and submit all required documents accurately.</li><li><strong>Selection Process:</strong> Shortlisted candidates might be invited for interviews which are conducted virtually most of the time. The selection process varies by program but often assesses academic prowess, motivation, and cross-cultural readiness.</li><li><strong>Notification:</strong> Successful candidates receive an admission offer along with information about the scholarship, funding details, and any additional steps to secure their place in the program.</li></ol><h4>Life as an Erasmus Mundus Scholar</h4><p>Being an Erasmus Mundus scholar is an exhilarating journey that offers a blend of academic rigors and unforgettable experiences.</p><ol><li><strong>Academic Adventures:</strong> Scholars engage in rigorous coursework, research projects, and seminars that challenge and broaden their intellectual horizons. Collaborating with peers from diverse backgrounds enhances the learning process.</li><li><strong>Cultural Immersion:</strong> Living in multiple countries exposes scholars to different cultures, languages, and ways of life. This immersion fosters personal growth, adaptability, and intercultural understanding.</li><li><strong>Global Connections:</strong> The friendships formed during this journey often span continents, creating a network of like-minded individuals with a shared passion for learning and exploration.</li><li><strong>Personal Development:</strong> Navigating new environments, dealing with challenges, and embracing change contribute to personal development and resilience. Scholars return home with a stronger sense of self and the world around them.</li></ol><h4>Conclusion</h4><p>I consider the Erasmus Mundus Scholarship as more than just financial support; it’s a transformative experience that shapes individuals like me into global citizens with a deep appreciation for diversity and a thirst for knowledge. Being an investment in education, culture, and personal growth, it yields lifelong rewards for aspiring scholars who dare to dream beyond borders and seek an extraordinary academic adventure. That being said, they should undoubtedly consider embarking on the Erasmus Mundus journey.</p><p><em>Thanks for reading! Do subscribe and share with everyone in your community and don’t forget to like and comment. Follow my Medium Page </em><a href="https://medium.com/u/6521c7f86bab"><em>Rishika Gupta</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7f0d78a48aa1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Time Series Forecasting with Python]]></title>
            <link>https://python.plainenglish.io/time-series-forecasting-with-python-17c9931ef1ca?source=rss-6521c7f86bab------2</link>
            <guid isPermaLink="false">https://medium.com/p/17c9931ef1ca</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[statistical-analysis]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[timeseries]]></category>
            <dc:creator><![CDATA[Rishika Gupta]]></dc:creator>
            <pubDate>Thu, 13 Jul 2023 17:04:11 GMT</pubDate>
            <atom:updated>2023-07-17T02:23:06.108Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*7rl4t4cnfETa6uwE" /><figcaption>Photo by <a href="https://unsplash.com/@aronvisuals?utm_source=medium&amp;utm_medium=referral">Aron Visuals</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Time series analysis involves techniques to examine time-based data to extract significant statistics and other notable features. Time series forecasting, on the other hand, consists in employing a model to make predictions about forthcoming values based on past observations.</p><p>Time series data finds extensive application in scenarios involving non-stationary data, such as economic indicators, weather patterns, stock prices, retail sales, etc. We will first analyse a time-series dataset and then forecast in this article. The dataset we will be using is carbon dioxide (an inbuilt dataset in statsmodels library of Python) which expresses the concentration in ppmv every week, which can be found <a href="https://www.statsmodels.org/dev/datasets/generated/co2.html">here</a>. Let’s commence!</p><h4>What is time series?</h4><ul><li>A <strong>time-series</strong> data is a series of data points or observations recorded at different or regular time intervals. In general, a time series is a sequence of data points taken at equally spaced time intervals. The frequency of recorded data points may be hourly, daily, weekly, monthly, quarterly or annually.</li><li><strong>Time-Series Forecasting</strong> is the process of using a statistical model to predict future values of a time series based on past results.</li><li>A time series analysis encompasses statistical methods for analyzing time series data. These methods enable us to extract meaningful statistics, patterns and other characteristics of the data. Time series are visualized with the help of line charts. So, time series analysis involves understanding the inherent aspects of the time series data so that we can create meaningful and accurate forecasts.</li><li>Applications of time series are used in statistics, finance or business applications. A very common example of time series data is the daily closing value of the stock index like NASDAQ or Dow Jones. Other common applications of time series are sales and demand forecasting, weather forecasting, econometrics, signal processing, pattern recognition and earthquake prediction.</li></ul><h4>What are the components of time series?</h4><ul><li><strong>Trend</strong> — The trend shows a general direction of the time series data over a long period of time. A trend can be increasing(upward), decreasing(downward), or horizontal(stationary).</li><li><strong>Seasonality</strong> — The seasonality component exhibits a trend that repeats with respect to timing, direction, and magnitude. Some examples include an increase in water consumption in summer due to hot weather conditions.</li><li><strong>Cyclical Component</strong> — These are the trends with no set repetition over a particular period of time. A cycle refers to the period of ups and downs, booms and slums of a time series, mostly observed in business cycles. These cycles do not exhibit a seasonal variation but generally occur over a time period of 3 to 12 years depending on the nature of the time series.</li><li><strong>Irregular Variation</strong> — These are the fluctuations in the time series data which become evident when trend and cyclical variations are removed. These variations are unpredictable, erratic, and may or may not be random.</li><li><strong>ETS Decomposition</strong> — ETS Decomposition is used to separate different components of a time series. The term ETS stands for Error, Trend and Seasonality.</li></ul><p>In this notebook, I conduct a time series analysis of CO2 concentration over time. Let’s now jump to the implementation.</p><h4>Step1 - Importing and Setting up</h4><p>First and foremost, we shall be importing the necessary Python modules and libraries.</p><pre># Importing the necessary libraries<br>import warnings<br>import itertools<br>import numpy as np<br>import matplotlib.pyplot as plt<br>import pandas as pd<br>import statsmodels.api as sm<br>import matplotlib<br>from pylab import rcParams</pre><p>To have fixed parameters throughout the program, we shall be fixing them beforehand, providing some consistency.</p><pre># Configure some parameters<br>warnings.filterwarnings(&quot;ignore&quot;)<br>plt.style.use(&#39;fivethirtyeight&#39;)<br>matplotlib.rcParams[&#39;axes.labelsize&#39;] = 10<br>matplotlib.rcParams[&#39;xtick.labelsize&#39;] = 8<br>matplotlib.rcParams[&#39;ytick.labelsize&#39;] = 8<br>matplotlib.rcParams[&#39;text.color&#39;] = &#39;k&#39;<br>rcParams[&#39;figure.figsize&#39;] = 18,8</pre><p>Next, as discussed in the initial section, let’s import the data from statsmodels of the Python library.</p><pre># Loading the data<br>df = sm.datasets.co2.load_pandas().data<br>df.head(5)</pre><h4>Step2 - Data Exploration</h4><p>Before we can jump into time-series modelling, it’s best to get an idea of the structure and ranges. This will also allow us to look for outliers or missing values that need to be corrected.</p><p>Since we are dealing with temporal data, we need to understand the range within which our data lies, so let’s check the minimum and maximum data points. In addition to this, we shall also check the length of the entire data.</p><pre># Checking for the min and max values<br><br>minDate = df.index.min()<br>maxDate = df.index.max()<br>print(f&#39;We have data starting from {minDate} until {maxDate}&#39;)</pre><p>This results in getting an output that states — “We have data starting from 1958–03–29 00:00:00 until 2001–12–29 00:00:00”.</p><pre># Checking the size of data<br>print(f&#39;Length of data is {len(df)}&#39;)</pre><p>This results in the length of data as 2284. Thus, all in all, we have data for almost 50 years. Further, in the case of time-series data, the dates should be used as an index. Since we imported the data from statsmodels it is set, otherwise we can set it using df.set_index(&#39;date&#39;).</p><pre># Checking the index of the data<br>print(df.index)</pre><pre>Output: <br>DatetimeIndex([&#39;1958-03-29&#39;, &#39;1958-04-05&#39;, &#39;1958-04-12&#39;, &#39;1958-04-19&#39;,<br>               &#39;1958-04-26&#39;, &#39;1958-05-03&#39;, &#39;1958-05-10&#39;, &#39;1958-05-17&#39;,<br>               &#39;1958-05-24&#39;, &#39;1958-05-31&#39;,<br>               ...<br>               &#39;2001-10-27&#39;, &#39;2001-11-03&#39;, &#39;2001-11-10&#39;, &#39;2001-11-17&#39;,<br>               &#39;2001-11-24&#39;, &#39;2001-12-01&#39;, &#39;2001-12-08&#39;, &#39;2001-12-15&#39;,<br>               &#39;2001-12-22&#39;, &#39;2001-12-29&#39;],<br>              dtype=&#39;datetime64[ns]&#39;, length=2284, freq=&#39;W-SAT&#39;)</pre><p>The dtype = datetime[ns] field confirms that our index is made of date stamp objects, while length = 2284 and freq = ‘W-SAT’ tells us that we have 2284 weekly date stamps starting on Saturdays.</p><p>Weekly data can be tricky to work with, so let’s use the monthly averages of our time series instead. This can be obtained by using the convenient resample function, which allows us to group the time series into buckets (1 month), apply a function on each group (mean), and combine the result (one row per group).</p><pre># Performing resampling of data from weekly to monthly<br>df = df[&#39;co2&#39;].resample(&#39;MS&#39;).mean()<br>df[:5]</pre><p>Here, the MS means that we group the data in buckets by month and ensure that we are using the start of each month as the timestamp.</p><pre># Verifying the data changed from weekly to monthly<br>print(df.index)</pre><pre>DatetimeIndex([&#39;1958-03-01&#39;, &#39;1958-04-01&#39;, &#39;1958-05-01&#39;, &#39;1958-06-01&#39;,<br>               &#39;1958-07-01&#39;, &#39;1958-08-01&#39;, &#39;1958-09-01&#39;, &#39;1958-10-01&#39;,<br>               &#39;1958-11-01&#39;, &#39;1958-12-01&#39;,<br>               ...<br>               &#39;2001-03-01&#39;, &#39;2001-04-01&#39;, &#39;2001-05-01&#39;, &#39;2001-06-01&#39;,<br>               &#39;2001-07-01&#39;, &#39;2001-08-01&#39;, &#39;2001-09-01&#39;, &#39;2001-10-01&#39;,<br>               &#39;2001-11-01&#39;, &#39;2001-12-01&#39;],<br>              dtype=&#39;datetime64[ns]&#39;, length=526, freq=&#39;MS&#39;)</pre><p>Now when we verify, the dtype = datetime[ns] field confirms that our index is made of date stamp objects, while length = 526 and freq = ‘MS’ tells us that we have 525 monthly date stamps.</p><p>Next, check the null values. This output tells us that there are 5 months with missing values in our time series. Generally, we should ‘fill in’ missing values if they are not too numerous so that we don’t have gaps in the data using the fillna() command. For simplicity, we can fill in missing values with the closest non-null value in our time series, although it is important to note that a rolling mean would sometimes be preferable.</p><pre># Checking for null values<br>print(df.isnull().sum())<br><br># Filling na values<br>df = df.fillna(df.bfill())<br>print(df.isnull().sum())</pre><p>Finally, visualizing the raw data to understand the pattern of the data.</p><pre># Plotting the data<br>df.plot(figsize=(8, 5))<br>plt.show()</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/741/1*p1O2nYVMwxU_Zh6sYrYntA.png" /><figcaption>CO2 data</figcaption></figure><p>Generally, for time-series data, there is a distinguishable pattern that appears when we plot the data. For example for sales data, the sales are always low at the beginning of the year and high at the end of the year, thus giving an upward trend within any single year.</p><p>So, to understand the patterns, we visualize our data using a method called time-series decomposition that allows us to decompose our time series into three distinct components: trend, seasonality, and noise.</p><pre>decomposition = sm.tsa.seasonal_decompose(df, model=&#39;additive&#39;)<br>fig = decomposition.plot()<br>plt.show()</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gfXAD_3hyk4hL-MVubp0HQ.png" /><figcaption>Decomposition of CO2 data into trend, seasonality and noise</figcaption></figure><p>The plot above clearly shows that the level of carbon dioxide is pretty stable.</p><h4>Step3 - Time Series Forecasting</h4><p>ARIMA stands for Autoregressive Integrated Moving Average. The ARIMA models are denoted with the notation ARIMA(p, d, q), wherein the parameters represent seasonality, trend and noise in the data.</p><p>First, we shall configure the parameters of p, q and d. In order to do that, we shall be searching for an optimal set of parameters in our defined grid, which shall yield the best performance for our model.</p><pre># Configuring the ARIMA model<br>p = d = q = range(0, 3)<br>pdq = list(itertools.product(p, d, q))<br>seasonal_pdq = [(x[0], x[1], x[2], 12) for x in list(itertools.product(p, d, q))]<br>print(&#39;Examples of parameter combinations for Seasonal ARIMA...&#39;)<br>print(&#39;SARIMAX: {} x {}&#39;.format(pdq[1], seasonal_pdq[1]))<br>print(&#39;SARIMAX: {} x {}&#39;.format(pdq[1], seasonal_pdq[2]))<br>print(&#39;SARIMAX: {} x {}&#39;.format(pdq[2], seasonal_pdq[3]))<br>print(&#39;SARIMAX: {} x {}&#39;.format(pdq[2], seasonal_pdq[4]))</pre><pre># Creating a grid-search for params<br>for param in pdq:<br>    for param_seasonal in seasonal_pdq:<br>        try:<br>            mod = sm.tsa.statespace.SARIMAX(df,<br>                                            order = param,<br>                                            seasonal_order = param_seasonal,<br>                                            enforce_stationarity = False,<br>                                            enforce_invertibility = False)<br><br>            results = mod.fit()<br><br>            print(&#39;ARIMA{}x{}12 - AIC:{}&#39;.format(param, param_seasonal, results.aic))<br>        except:<br>            continue</pre><p>The output from this code suggests that SARIMAX(2, 0, 1)x(1, 0, 2, 12) yields the lowest AIC value of 271.77. Therefore we should consider this to be the optimal option.</p><p>Next, we can fit the ARIMA model with the optimal parameters found above and plot the diagnostics to understand any unusual behaviour.</p><pre># Fitting the ARIMA model to the optimal parameters found in the above step<br>mod = sm.tsa.statespace.SARIMAX(df,<br>                                order = (2, 0, 1),<br>                                seasonal_order = (1, 0, 2, 12),<br>                                enforce_stationarity = False,<br>                                enforce_invertibility = False)<br><br>results = mod.fit()<br><br>print(results.summary().tables[1])<br><br># Plotting the diagonstics to identify any unusual behaviour<br>results.plot_diagnostics(figsize=(16, 8))<br>plt.show()</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xfJMfdNxRUayrUlAGUn0YA.png" /><figcaption>Diagnostic Summary for CO2 data</figcaption></figure><p>It is not perfect since our model diagnostics suggest that the model residuals are unstable.</p><p>Next, we shall be validating the forecasts of our ARIMA model. To help us understand the accuracy of our forecasts, we compare predicted sales to real CO2 of the time series, and we set forecasts to start at 1997–01–01 to the end of the data.</p><pre># Running model on validation data<br>pred = results.get_prediction(start = pd.to_datetime(&#39;1997-01-01&#39;), dynamic = False)<br>pred_ci = pred.conf_int()<br>ax = df[&#39;1958&#39;:].plot(label = &#39;observed&#39;)<br>pred.predicted_mean.plot(ax = ax, label = &#39;One-step ahead Forecast&#39;, alpha = .7, figsize = (14, 7))<br>ax.fill_between(pred_ci.index,<br>                pred_ci.iloc[:, 0],<br>                pred_ci.iloc[:, 1], color = &#39;k&#39;, alpha = .2)<br>ax.set_xlabel(&#39;Date&#39;)<br>ax.set_ylabel(&#39;CO2 Level&#39;)<br>plt.legend()<br>plt.show()</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Xbbj1G1foh2MnS1Scyeyxw.png" /><figcaption>Validating forecasts of ARIMA model</figcaption></figure><pre>df_forecasted = pred.predicted_mean<br>df_truth = df[&#39;1997-01-01&#39;:]<br>mse = ((df_forecasted - df_truth) ** 2).mean()<br>print(&#39;The Mean Squared Error of our forecasts is {}\n&#39;.format(round(mse, 2)))<br>print(&#39;The Root Mean Squared Error of our forecasts is {}&#39;.format(round(np.sqrt(mse), 2)))</pre><p>The Mean Squared Error of our forecasts is 0.1 while the Root Mean Squared Error of our forecasts is 0.31</p><p>Well, it is obvious that given the energy consumption trends followed by people, the CO2 levels are increasing. Since, the MSE and RMSE are very small, in my opinion, the time-series model can be said here in this case to be more than good.</p><p>Next, we shall be producing and visualizing forecasts.</p><pre>pred_uc = results.get_forecast(steps = 250)<br>pred_ci = pred_uc.conf_int()<br>ax = df.plot(label=&#39;observed&#39;, figsize=(14, 7))<br>pred_uc.predicted_mean.plot(ax=ax, label = &#39;Forecast&#39;)<br>ax.fill_between(pred_ci.index,<br>                pred_ci.iloc[:, 0],<br>                pred_ci.iloc[:, 1], color = &#39;k&#39;, alpha = .25)<br>ax.set_xlabel(&#39;Date&#39;)<br>ax.set_ylabel(&#39;CO2 Level&#39;)<br>plt.legend()<br>plt.show()</pre><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2M0ga3sW22BtykNrCAI5gg.png" /><figcaption>Producing and visualizing forecasts</figcaption></figure><p>Our model clearly captured CO2 trends. As we forecast further out into the future, it is natural for us to become less confident in our values. This is reflected by the confidence intervals generated by our model, which grow larger as we move further out into the future.</p><p>You can find the entire source code on my <a href="https://github.com/risg99/Time-Series-Analysis/blob/main/time_series_forecasting.ipynb">GitHub</a>. I look forward to hearing feedback or questions.</p><p><em>Thanks for reading! Do subscribe and share with everyone in your community and don’t forget to like and comment. Follow my Medium Page </em><a href="https://medium.com/u/6521c7f86bab"><em>Rishika Gupta</em></a><em>.</em></p><p><em>More content at </em><a href="https://plainenglish.io/"><strong><em>PlainEnglish.io</em></strong></a><em>.</em></p><p><em>Sign up for our </em><a href="http://newsletter.plainenglish.io/"><strong><em>free weekly newsletter</em></strong></a><em>. Follow us on </em><a href="https://twitter.com/inPlainEngHQ"><strong><em>Twitter</em></strong></a>, <a href="https://www.linkedin.com/company/inplainenglish/"><strong><em>LinkedIn</em></strong></a><em>, </em><a href="https://www.youtube.com/channel/UCtipWUghju290NWcn8jhyAw"><strong><em>YouTube</em></strong></a><em>, and </em><a href="https://discord.gg/GtDtUAvyhW"><strong><em>Discord</em></strong></a><strong><em>.</em></strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=17c9931ef1ca" width="1" height="1" alt=""><hr><p><a href="https://python.plainenglish.io/time-series-forecasting-with-python-17c9931ef1ca">Time Series Forecasting with Python</a> was originally published in <a href="https://python.plainenglish.io/">Python in Plain English</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>