<?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 Garima Gupta on Medium]]></title>
        <description><![CDATA[Stories by Garima Gupta on Medium]]></description>
        <link>https://medium.com/@ggupta87?source=rss-89c556addf6e------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*zEBP7t0WIimdMkgFV1dz7g.jpeg</url>
            <title>Stories by Garima Gupta on Medium</title>
            <link>https://medium.com/@ggupta87?source=rss-89c556addf6e------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 21:12:36 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@ggupta87/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[“Doer” or “Decision-Maker” : Choose your own adventure]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-snippet">Have you ever felt like you&#x2019;re just working with numbers in your data job, but not really making a difference? You&#x2019;re not the only one&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@ggupta87/doer-or-decision-maker-choose-your-own-adventure-4073536e3af2?source=rss-89c556addf6e------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@ggupta87/doer-or-decision-maker-choose-your-own-adventure-4073536e3af2?source=rss-89c556addf6e------2</link>
            <guid isPermaLink="false">https://medium.com/p/4073536e3af2</guid>
            <dc:creator><![CDATA[Garima Gupta]]></dc:creator>
            <pubDate>Thu, 18 Jan 2024 18:17:45 GMT</pubDate>
            <atom:updated>2024-01-18T18:17:45.688Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Motherhood Unveiled: Navigating the Challenges and Celebrating the Joys]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-snippet">Not too long ago, I openly acknowledged that career and personal advancement often represent an uneven playing field for women. However&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@ggupta87/motherhood-unveiled-navigating-the-challenges-and-celebrating-the-joys-7b9a12d5619a?source=rss-89c556addf6e------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@ggupta87/motherhood-unveiled-navigating-the-challenges-and-celebrating-the-joys-7b9a12d5619a?source=rss-89c556addf6e------2</link>
            <guid isPermaLink="false">https://medium.com/p/7b9a12d5619a</guid>
            <dc:creator><![CDATA[Garima Gupta]]></dc:creator>
            <pubDate>Fri, 15 Sep 2023 06:12:42 GMT</pubDate>
            <atom:updated>2023-09-15T06:12:42.382Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Do’s & Don’ts of A/B Tests]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@ggupta87/dos-don-ts-of-a-b-tests-86240fd00fce?source=rss-89c556addf6e------2"><img src="https://cdn-images-1.medium.com/max/1420/1*ERxReMJsCc4izZyZNeoZzA.png" width="1420"></a></p><p class="medium-feed-snippet">A/B test, also known as &#x201C;Hypothesis test&#x201D; or &#x201C;Split test&#x201D; is a controlled experiment, used to test if there is a real impact of any change&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@ggupta87/dos-don-ts-of-a-b-tests-86240fd00fce?source=rss-89c556addf6e------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@ggupta87/dos-don-ts-of-a-b-tests-86240fd00fce?source=rss-89c556addf6e------2</link>
            <guid isPermaLink="false">https://medium.com/p/86240fd00fce</guid>
            <category><![CDATA[ab-test]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[learning]]></category>
            <dc:creator><![CDATA[Garima Gupta]]></dc:creator>
            <pubDate>Mon, 07 Mar 2022 04:25:10 GMT</pubDate>
            <atom:updated>2022-03-07T04:25:10.590Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[How cross-domain expertise can help you as a Data Scientist]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@ggupta87/how-cross-domain-expertise-can-help-you-as-a-data-scientist-eda99c70f196?source=rss-89c556addf6e------2"><img src="https://cdn-images-1.medium.com/max/1622/1*BILflxXLLLYdqVSGXa-h6A.png" width="1622"></a></p><p class="medium-feed-snippet">Did you get a job offer in a completely unknown domain? Don&#x2019;t worry! Learn the art of identifying and applying cross-domain skills and&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@ggupta87/how-cross-domain-expertise-can-help-you-as-a-data-scientist-eda99c70f196?source=rss-89c556addf6e------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@ggupta87/how-cross-domain-expertise-can-help-you-as-a-data-scientist-eda99c70f196?source=rss-89c556addf6e------2</link>
            <guid isPermaLink="false">https://medium.com/p/eda99c70f196</guid>
            <category><![CDATA[ml-so-good]]></category>
            <category><![CDATA[cross-domain]]></category>
            <category><![CDATA[data-science-leadership]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[Garima Gupta]]></dc:creator>
            <pubDate>Tue, 25 Jan 2022 17:11:45 GMT</pubDate>
            <atom:updated>2022-01-26T13:00:48.188Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Why do ML models fail in Production?]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@ggupta87/why-do-ml-models-fail-in-production-505d2cf58d5a?source=rss-89c556addf6e------2"><img src="https://cdn-images-1.medium.com/max/1058/1*-2I5tKnoktYaJHo5jInT_A.png" width="1058"></a></p><p class="medium-feed-snippet">Read if you think Machine learning has magical powers.</p><p class="medium-feed-link"><a href="https://medium.com/@ggupta87/why-do-ml-models-fail-in-production-505d2cf58d5a?source=rss-89c556addf6e------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@ggupta87/why-do-ml-models-fail-in-production-505d2cf58d5a?source=rss-89c556addf6e------2</link>
            <guid isPermaLink="false">https://medium.com/p/505d2cf58d5a</guid>
            <category><![CDATA[ml-model]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[Garima Gupta]]></dc:creator>
            <pubDate>Mon, 10 Jan 2022 00:20:05 GMT</pubDate>
            <atom:updated>2022-01-10T00:20:05.651Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Leadership: One person, many roles]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@ggupta87/leadership-one-person-many-roles-e32df62562a?source=rss-89c556addf6e------2"><img src="https://cdn-images-1.medium.com/max/1948/1*HXXwBv2RQ41ZXlN8KQAXWQ.png" width="1948"></a></p><p class="medium-feed-snippet">I have had several opportunities to experience and practice leadership in many companies and with different teams, where people differed&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@ggupta87/leadership-one-person-many-roles-e32df62562a?source=rss-89c556addf6e------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@ggupta87/leadership-one-person-many-roles-e32df62562a?source=rss-89c556addf6e------2</link>
            <guid isPermaLink="false">https://medium.com/p/e32df62562a</guid>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[feedback]]></category>
            <category><![CDATA[leadership-skills]]></category>
            <category><![CDATA[mentoring]]></category>
            <category><![CDATA[management-and-leadership]]></category>
            <dc:creator><![CDATA[Garima Gupta]]></dc:creator>
            <pubDate>Sun, 19 Dec 2021 00:29:24 GMT</pubDate>
            <atom:updated>2021-12-19T00:29:24.451Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Python Pandas : Selecting the data]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@ggupta87/python-pandas-selecting-the-data-541d9087bec5?source=rss-89c556addf6e------2"><img src="https://cdn-images-1.medium.com/max/1944/1*eFqI3iG8_950OQUsC7D5Hg.png" width="1944"></a></p><p class="medium-feed-snippet">One of the main component to explore any data or build any Machine learning model is to have the ability to select and subset data and&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@ggupta87/python-pandas-selecting-the-data-541d9087bec5?source=rss-89c556addf6e------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@ggupta87/python-pandas-selecting-the-data-541d9087bec5?source=rss-89c556addf6e------2</link>
            <guid isPermaLink="false">https://medium.com/p/541d9087bec5</guid>
            <category><![CDATA[python]]></category>
            <category><![CDATA[subsets]]></category>
            <category><![CDATA[data]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[pandas]]></category>
            <dc:creator><![CDATA[Garima Gupta]]></dc:creator>
            <pubDate>Sat, 30 May 2020 23:50:49 GMT</pubDate>
            <atom:updated>2020-05-30T23:50:49.282Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Most commonly used Pandas functions to understand your dataset]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@ggupta87/most-commonly-used-pandas-functions-to-understand-your-dataset-526a0e5bd4d?source=rss-89c556addf6e------2"><img src="https://cdn-images-1.medium.com/max/1842/1*WbdjvyBlIkJ-Jb8cZqjCwA.png" width="1842"></a></p><p class="medium-feed-snippet">Before building and deploying any machine learning model, it is extremely important to understand your data and pre-process it before&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@ggupta87/most-commonly-used-pandas-functions-to-understand-your-dataset-526a0e5bd4d?source=rss-89c556addf6e------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@ggupta87/most-commonly-used-pandas-functions-to-understand-your-dataset-526a0e5bd4d?source=rss-89c556addf6e------2</link>
            <guid isPermaLink="false">https://medium.com/p/526a0e5bd4d</guid>
            <category><![CDATA[data-cleaning]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[data-analysis]]></category>
            <category><![CDATA[pandas]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[Garima Gupta]]></dc:creator>
            <pubDate>Sat, 30 May 2020 22:57:22 GMT</pubDate>
            <atom:updated>2020-05-30T22:57:22.539Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Exploratory Data Analysis with COVID-19 data]]></title>
            <link>https://medium.com/@ggupta87/exploratory-data-analysis-with-covid-19-data-india-189300ed82bb?source=rss-89c556addf6e------2</link>
            <guid isPermaLink="false">https://medium.com/p/189300ed82bb</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[exploratory-data-analysis]]></category>
            <category><![CDATA[covid19]]></category>
            <category><![CDATA[eda]]></category>
            <dc:creator><![CDATA[Garima Gupta]]></dc:creator>
            <pubDate>Sun, 29 Mar 2020 19:15:15 GMT</pubDate>
            <atom:updated>2020-03-29T19:15:15.178Z</atom:updated>
            <content:encoded><![CDATA[<p>Exploratory data analysis, also known as <em>Data Exploration, </em>is a tool to understand a dataset better through summarizing, visualizing and becoming familiar with the important characteristics of a data set. EDA is not only a critical step in building any Machine Learning model but a useful way to generate insights from the data.</p><h3>Why EDA is important</h3><p>Any Machine learning model follows the principle of <em>garbage-in, garbage-out.</em>This means if wrong data (skewed, outliers, too many missing values) is fed to a model, it will lead to inaccurate models and hence bad and incorrect outcomes.</p><ul><li>Some of the key benefits of performing EDA on the data sets are:</li><li>Determine relationship (or lack of) among variables</li><li>Dimensionality reduction</li><li>Feature selection</li><li>Generating insights</li><li>Building and conforming hypothesis</li></ul><h3>Methods of EDA</h3><p>Based on the number of variables explored and whether the medium of exploration is a graph, there are 4 methods of performing EDA —</p><ul><li><strong>Univariate non-graphical </strong>: Exploring single variable using descriptive statistics</li><li><strong>Univariate graphical</strong> : Exploring single variable using different plots</li><li><strong>Multi-variate non-graphical</strong> : Exploring multiple variables using descriptive statistics and matrices</li><li><strong>Multi-variate graphical</strong> : Exploring multiple variables using plots</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/860/1*l2xooivtB3ghe4bTo7roJg.png" /></figure><h3>Case Study</h3><p>With current COVID-19 pandemic and lot of unknowns and uncertainties associated with it, data is one of the most power tool available to understand the problem and combat it by uncovering hidden information and pattern.</p><p>My family lives in India and with so much of misinformation floating around, I decided to explore available COVID dataset for India and understand country’s state. Using univariate and multivariate exploratory data analysis techniques, I was able to generate initial-level insights as covered below.</p><h4><strong>Available datasets</strong></h4><ul><li>Daily cases count</li><li>Census data</li><li>Hospital beds in each Indian states</li><li>Individual level data for positive cases</li></ul><h4>How much data is missing</h4><ul><li><strong>Daily cases count</strong>: The last few rows of the data set are completely bank, as can be seen in the missingness plot below.This could be due to blank rows in csv files while uploading the data</li><li><strong>Census data</strong>: No missing rows or columns or cells found</li><li><strong>Hospital beds data</strong>: Besides few rows missing in one of the column, last two unknown/unnamed columns are completely blank</li><li><strong>Individual level data for positive cases</strong>: Multiple columns and rows are missing in this data set, which is not surprising as it is personally identifiable data, hence expected to be missing</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/849/1*tOGORny4z_QeNjxXZW9fEQ.png" /><figcaption>Missingness plot for available India COVID-19 datasets</figcaption></figure><h4><strong>Confirmed cases growth</strong></h4><p>Exponential curve continues for India for positive COVID-19 cases, while deaths have not peaked yet. Strict proactive actions now can help <em>flattening the curve</em> sooner</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/319/1*saLjcyYjTtj2WRtBdOP5EQ.png" /></figure><h4><strong>How are different Indian states doing</strong></h4><h4>1. Confirmed cases with time</h4><ul><li><strong>“Kerala” &amp; “Maharashtra” are impacted worst, </strong>contributing to 38% of total positive cases in India</li><li><strong>Foreign tourism/travel may be contributing most to state of Haryana </strong>— possibly due to closeness to Delhi airport (gateway to international travel)</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/475/1*ANIul3z_WMHSklG9Z0ZiEQ.png" /></figure><h4>2. Confirmed cases vs Population density</h4><ul><li><strong>“Kerala” &amp; “Maharashtra” are </strong>not effective in social-distancing — top two states with highest number of confirmed cases and having lowest densities</li><li><strong>Delhi</strong> has been effective in implementing social-distancing measures — highest population density but 7th among states in positive cases</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/489/1*9-YhoyLz4ajvkTht7tosYQ.png" /></figure><h4>3. Confirmed cases vs available Hospital beds</h4><ul><li><strong>Only 0.12% hospital beds are available </strong>for total population of India</li><li>States like <strong>Kerala, Maharashtra, UP, Rajasthan, Karnataka and Delhi need to tighten social distancing </strong>measures due to higher number of confirmed cases at initial stage and low availability of beds</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/489/1*9HYfcLrG69VA4oDhHHz0Mw.png" /></figure><h3>Conclusion</h3><p>EDA helped me to understand ongoing COVID-19 situation in India while sitting in United States by performing univariate and multi-variate analysis of the available dataset. Such is the power of this tool. While Machine learning models are effective in learning patterns and making predictions but they are only as accurate as the data fed into these models. Hence, ensuring the quality of the training data by performing EDA is vital for the performance of any ML model.</p><p><em>Sources :</em></p><p><a href="https://www.kaggle.com/sudalairajkumar/covid19-in-india"><em>COVID-19 in India dataset by Kaggle</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=189300ed82bb" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>