<?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 Sarannithish K S on Medium]]></title>
        <description><![CDATA[Stories by Sarannithish K S on Medium]]></description>
        <link>https://medium.com/@saran_iceapple?source=rss-8ed1cba6880------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*BiC0w00MyBHzC4K2</url>
            <title>Stories by Sarannithish K S on Medium</title>
            <link>https://medium.com/@saran_iceapple?source=rss-8ed1cba6880------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 17:33:41 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@saran_iceapple/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[The Path to AI]]></title>
            <link>https://medium.com/iceapple-tech-talks/the-path-to-ai-9d40c9d3fd9d?source=rss-8ed1cba6880------2</link>
            <guid isPermaLink="false">https://medium.com/p/9d40c9d3fd9d</guid>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Sarannithish K S]]></dc:creator>
            <pubDate>Sun, 15 Oct 2023 06:18:01 GMT</pubDate>
            <atom:updated>2023-10-19T08:20:59.013Z</atom:updated>
            <content:encoded><![CDATA[<p>With all the gold rush happening now on the AI , most of us would want to jump into it and make hay while the sun shines.</p><p>This blog is focussed on guiding aspirants in the right path to becoming an AI engineer.</p><p><strong>Where to start ?</strong></p><p>The cold truth is that, it cannot be understood like any new software framework or library. The common mistake which people do is to start at the wrong place.</p><p>We normally get carried away by the fancy stuff going around and start experimenting or searching solutions for bigger problems without understanding about the smaller moving parts . We start building an aeroplane without understanding Bernoulli’s equation !!</p><p>The first step in becoming an AI engineer is to get stronger with your math, especially calculus. Then comes other siblings like probabilities, statistics ,etc..</p><p><strong>How long does it take ?</strong></p><p>Getting to the depth of the basic concepts and gaining the ability to apply them in solving real-time problems takes anywhere from 6 months to a year.</p><p>Don’t get pessimistic by the timeline, it is of course the time taken by me and people can get to it faster than that.</p><p>At an average , one may need to spend at least 20 hours a week to get to the crux of AI.</p><p><strong>The AI Umbrella</strong></p><p>AI is a very big word which encompasses lots of subjects under it. One cannot master AI as a whole and it is a continuous learning process.<br>I am listing the branches below. Of course this is not entirety</p><p><strong>Problem Based</strong></p><ol><li>Natural Language Processing</li></ol><p>2. Computer Vision</p><p>3. Generative AI</p><p><strong>Technique Based</strong></p><ol><li>Machine Learning</li></ol><p>2. Deep Learning</p><p><strong>Data Based</strong></p><p>1.Supervised Learning</p><p>2.Unsupervised Learning</p><p>3.Weakly Supervised Learning</p><p><strong>The AI Stack</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/695/1*JuWg5bFsksulqm1EGpZ7Fg.png" /><figcaption>The AI Stack</figcaption></figure><p><strong>Learning map</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*R7jCBQkYPagJmrxw4BjyqQ.png" /><figcaption>AI Learning Map</figcaption></figure><p>The blocks marked in “cyan” are the major categories ( subjects ) and the ones marked in “grey” are the sub-categories ( Topics ).</p><p>The learning path remains common for all AI engineers until they know the basic math, programming and data analysis techniques.</p><p>The path can then be diversified based on the interest if the engineer.</p><p><strong>Some places where you can start</strong></p><p><a href="https://www.coursera.org/collections/best-machine-learning-ai">https://www.coursera.org/collections/best-machine-learning-ai</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9d40c9d3fd9d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/iceapple-tech-talks/the-path-to-ai-9d40c9d3fd9d">The Path to AI</a> was originally published in <a href="https://medium.com/iceapple-tech-talks">IceApple DevLogs</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Confusion to Clarity]]></title>
            <link>https://medium.com/iceapple-tech-talks/confusion-to-clarity-94c8fdb12224?source=rss-8ed1cba6880------2</link>
            <guid isPermaLink="false">https://medium.com/p/94c8fdb12224</guid>
            <category><![CDATA[confusion-matrix]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[Sarannithish K S]]></dc:creator>
            <pubDate>Sun, 24 Sep 2023 10:09:01 GMT</pubDate>
            <atom:updated>2023-09-25T14:17:31.715Z</atom:updated>
            <content:encoded><![CDATA[<h3>Confusion to Clarity ( Confusion matrix demystified )</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/768/1*YfcWY1UiS2ehiGeN1xWkIg.jpeg" /></figure><p>I would like to start this post by saying that “Confusion Matrix” really confused me too :) I really thought that it was the same for everyone , and hence the name “Confusion Matrix” , but really not.</p><p>The name was arrived in terms of the model being built, and used to gauge how much the machine learning model is confused with identifying the correct classes.</p><p><strong>Story:</strong></p><p>Somewhere in the future ( maybe 2100 AD) , in Eekaako , there was a medical screening centre which had state of the art AI systems to screen patients coming in. The screening centre was specialised in screening for a rare disease named “Eklakophobia” which broke out that year.</p><p>Although rare , the disease can spread extremely fast and infect the surroundings unless a strict quarantine protocol is followed ( Something like COVID ).</p><p><strong>AI entry:</strong></p><p>The job of the AI system is to screen the individuals and classify them as either positive or negative for “Eklakophobia”.</p><p>Since AI is also human, it can make mistakes and we have to be tolerant with it, but how much ?</p><p>There are 100 patients coming in for screening , out of which 60 are really infected with “Eklakophobia”, the remaining 40 are not. <br>So , the ground truth is</p><p><strong>Infected : 60 <br>Not Infected : 40</strong></p><p><a href="http://ai_error/">All the 100 patients are screened by the AI system and the results of the AI system are as follows</a></p><p><strong>Infected : 50<br>Not infected : 50</strong></p><p><strong>The confusion :<br></strong>Since this is a highly contagious disease, we cannot afford to leave even a single patient who is actually infected with the disease.</p><p>But the AI predictions are not matching with the ground truth and we need to identify how much we have deviated from the reality.</p><p>Let’s check the AI result of every instance ( every patient ) with the ground truth and put them into the below categories</p><p><strong>True Positive ( TP ) </strong>— Really infected , AI also predicts as infected</p><p><strong>True Negative (TN) </strong>— Really not infected , AI also predicts as not infected</p><p><strong>False Positive (FP)</strong> — Really not infected , AI predicts as infected i.e, AI has falsely predicted as positive for infection.</p><p><strong>False Negative (FN)</strong> — Really infected, AI predicts as not infected i.e, AI has falsely predicted as negative for infection.</p><p>Lets’s draw the confusion matrix</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/394/1*-6ozzkLBJphr35G-aCjOlw.jpeg" /></figure><p><strong>What these number say ?</strong></p><p>I think, now we get some clarity from the confusion…Uh, confusion matrix :)</p><p><strong>Infected<br></strong>Truth — 60 AI prediction — 50</p><p><strong>Not infected<br></strong>Truth — 40 AI prediction — 50</p><p>The ideal prediction should be <br>TP — 60 <br>TN — 40 <br>FP — 0<br>FN — 0<br>But from the confusion matrix , we see that there are 10 FPs ,which means that 10 not-infected persons were misclassified as infected.</p><p>There are 30 FNs which means that 30 people who were infected were misclassified as not-infected.</p><p>Oh God !! The issue is even more serious that we envisioned. <br>At a glance the initial numbers showed that we had only 10 misclassifications ( <a href="http://ai_error/">https://medium.com/r/?url=http%3A%2F%2FAI_Error</a> )</p><p>But confusion matrix revealed that there are 30 infected people who were screened as not infected. This could have been catastrophic….</p><p>Thanks to the confusion matrix for saving the day !!!!</p><p>Always we get to clarity from confusion, but we need to look into it deeply !!!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=94c8fdb12224" width="1" height="1" alt=""><hr><p><a href="https://medium.com/iceapple-tech-talks/confusion-to-clarity-94c8fdb12224">Confusion to Clarity</a> was originally published in <a href="https://medium.com/iceapple-tech-talks">IceApple DevLogs</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Source Control for Noobies]]></title>
            <link>https://medium.com/iceapple-tech-talks/source-control-for-noobies-a5edde8adf3a?source=rss-8ed1cba6880------2</link>
            <guid isPermaLink="false">https://medium.com/p/a5edde8adf3a</guid>
            <category><![CDATA[source-control]]></category>
            <category><![CDATA[backend-development]]></category>
            <category><![CDATA[git]]></category>
            <category><![CDATA[scm]]></category>
            <dc:creator><![CDATA[Sarannithish K S]]></dc:creator>
            <pubDate>Fri, 05 May 2023 14:24:55 GMT</pubDate>
            <atom:updated>2023-08-28T14:05:57.930Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*nR4Ch7Y7xcQMIYA5" /><figcaption>Photo by <a href="https://unsplash.com/fr/@mitchellleach?utm_source=medium&amp;utm_medium=referral">Mitchell Leach</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p><strong>What is Source Control ?</strong></p><p>Hey folks, what is source control? Is it controlling something at source?</p><p>Well, not exactly!! This is a beginners guide to understand the fundamentals of source control.</p><p>Let’s start with the problem statement first. <br> <br> Assuming only a single developer has coded a product. <br> The product has come to a good shape, and it is delivered to a customer.</p><p>Many features were added to the product and now become difficult to be maintained by a single developer and the company hires more people to maintain the product.</p><p>The new developer now comes in and looks at the code and wants to fix a bug.</p><p>The new developer has to understand the history of the bug to fix it.</p><p>What are the Options?<br> <br> 1. Ask the previous developer who has developed the code.</p><p>What if the previous developer has left the company?</p><p>It is going to be catastrophic where fixing one bug can introduce many more.</p><p>So we need a solution to know what all happened in the development life cycle, so that anyone can trace back the time and see what change introduced this bug.</p><p>That’s where control comes to the rescue. Source control system keeps track of the changes that undergo in a file by maintaining log of every difference the file undergoes.</p><p><strong>What is the business impact ?</strong></p><p>Let’s jump into one more story. Let’s assume the product is given to a customer (named Alice) and it is in use.</p><p>We have added more features on top of existing product which is not yet released to the Alice. Customer named Bob buys our product with the added set of features.</p><p>All of a sudden, Alice comes back with a bug in the product and wants it to be fixed, but she is not ready to pay the cost for the additional features.</p><p>What do we do now? Our product now has all the added features and if we give the fix on top of it, Alice will be able be able to use the new features.</p><p>Again, source control is there to save us.</p><p>Source control system helps version the product and maintain the releases as well.</p><p>Every single release can be independently tracked and delivered so that the business does not end up having a monolith which is difficult to sell.</p><p>With the help of versions and releases you will be able to provide the bug fix on top of the feature list which Alice already had without giving the additional features built for Bob.</p><p><strong>A very powerful tool for source control is Git.</strong></p><p>So, let’s jump into the details folks.</p><p><strong>Git architecture</strong></p><p>Git provides a two layered architecture for convenience and safety.</p><p>The first layer is remote and the second one is local.</p><p>The remote layer resides on a server which can be hosted in the cloud or any data centre. The local layer runs on the development machine.</p><p>Git tracks all the changes made to the files and provide mechanisms to freeze them.</p><p>The freezing of change needs to happen on both at the local layer and the remote layer.</p><p>The changes are initially frozen locally and then later pushed to the server.</p><p>This architecture provides an advantage to the developer to still hold lot of changes locally without having to push them or publish them until the changes are ready to be exposed to the world.</p><p>Repositories are the basic controllable/synchronizable entities and files are the basic trackable entities in the world of source control.</p><p>Repositories can contain many folders and files, but control and sync actions happen only at the repository level.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/501/1*9q77i1Zit6VrMbz6ooBKVw.jpeg" /></figure><p><strong>Git Branches</strong></p><p>Git branches are like independent copies of code which gives freedom to the developers to experiment all their stuff without impacting the product mainline.<br> <br> It also helps to keep the main product line intact without being exposed to any experimental code.</p><p><strong>Sample branching strategy</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/632/1*223ARq8A2KNygwLn55-bkA.jpeg" /></figure><p>The creation of “Feature” branch gives full freedom to the developer to play with his/her code without impacting the “Main” or “Dev” branch which are intended for customer and development release respectively.<br> <br> The changes in the feature branch can be merged to the appropriate release branches only after the feature has become stable.</p><h3><strong>Let’s watch it in action</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/721/1*9EXNsDqu5BmUOQ3UheX3yg.jpeg" /></figure><h3>Git Commands</h3><p><strong>git init</strong></p><p>· The command git init is used to create an empty Git repository.</p><p>· After the git init command is used, a .git folder is created in the directory with some subdirectories. Once the repository is initialized, the process of creating other files begins.</p><p><strong>git add</strong></p><p>· Add command is used after checking the status of the files, to add those files to the staging area.</p><p>· Before running the commit command, “git add” is used to add any new or modified files.</p><p><strong>git commit</strong></p><p>· The commit command makes sure that the changes are saved to the local repository.</p><p>· The command “git commit –m &lt;message&gt;” allows you to describe everyone and help them understand what has happened.</p><p><strong>git status</strong></p><p>· The git status command tells the current state of the repository.</p><p>· The command provides the current working branch. If the files are in the staging area, but not committed, it will be shown by the git status. Also, if there are no changes, it will show the message no changes to commit, working directory clean.</p><p><strong>git config</strong></p><p>· The git config command is used initially to configure the user.name and user.email. This specifies what email id and username will be used from a local repository.</p><p>· When git config is used with — global flag, it writes the settings to all repositories on the computer.</p><p><strong>git branch</strong></p><p>· The git branch command is used to determine what branch the local repository is on.</p><p>· The command enables adding and deleting a branch.</p><p>git checkout</p><p>· The git checkout command is used to switch branches, whenever the work is to be started on a different branch.</p><p>· The command works on three separate entities: files, commits, and branches.</p><p><strong>git merge</strong></p><p>· The git merge command is used to integrate the branches together. The command combines the changes from one branch to another branch.</p><p>· It is used to merge the changes in the staging branch to the stable branch.</p><p><strong>Git Commands: Working With Remote Repositories</strong></p><p><strong>git remote</strong></p><p>· The git remote command is used to create, view, and delete connections to other repositories.</p><p>· The connections here are not like direct links into other repositories, but as bookmarks that serve as convenient names to be used as a reference.</p><p><strong>git clone</strong></p><p>· The git clone command is used to create a local working copy of an existing remote repository.</p><p>· The command downloads the remote repository to the computer. It is equivalent to the Git init command when working with a remote repository.</p><p><strong>git pull</strong></p><p>· The git pull command is used to fetch and merge changes from the remote repository to the local repository.</p><p>· The command “git pull origin master” copies all the files from the master branch of the remote repository to the local repository.</p><p><strong>git fetch</strong></p><p>· The “<strong>git fetch</strong>” <strong>command</strong> is used to pull the updates from remote-tracking branches.</p><p>· Additionally, we can get the updates that have been pushed to our remote branches to our local machines.</p><p>· As we know, a branch is a variation of our repositories main code, so the remote-tracking branches are branches that have been set up to pull and push from remote repository.</p><p><strong>git push</strong></p><p>· The command <a href="https://www.simplilearn.com/tutorials/git-tutorial/git-push-command">git push</a> is used to transfer the commits or pushing the content from the local repository to the remote repository.</p><p>· The command is used after a local repository has been modified, and the modifications are to be shared with the remote team members.</p><p>Some Advanced Git Commands</p><p><strong>git stash</strong></p><p>· The git stash command takes your modified tracked files and saves it on a pile of incomplete changes that you can reapply at any time. To go back to work, you can use the stash pop.</p><p>· The git stash command will help a developer switch branches to work on something else without committing to incomplete work.</p><p><strong>git log</strong></p><p>· The git log command shows the order of the commit history for a repository.</p><p>· The command helps in understanding the state of the current branch by showing the commits that lead to this state.</p><p><strong>Git Commands Reference : </strong><a href="https://www.simplilearn.com/tutorials/git-tutorial/git-commands"><strong>https://www.simplilearn.com/tutorials/git-tutorial/git-commands</strong></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a5edde8adf3a" width="1" height="1" alt=""><hr><p><a href="https://medium.com/iceapple-tech-talks/source-control-for-noobies-a5edde8adf3a">Source Control for Noobies</a> was originally published in <a href="https://medium.com/iceapple-tech-talks">IceApple DevLogs</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>