<?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 Catherine Liang on Medium]]></title>
        <description><![CDATA[Stories by Catherine Liang on Medium]]></description>
        <link>https://medium.com/@dailyimprove2012?source=rss-a38c442737c0------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*S3ajLZN9kGUbBFSE.jpg</url>
            <title>Stories by Catherine Liang on Medium</title>
            <link>https://medium.com/@dailyimprove2012?source=rss-a38c442737c0------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 08:10:57 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@dailyimprove2012/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[Kaggle BIPOC Program Application Tips]]></title>
            <link>https://dailyimprove2012.medium.com/kaggle-bipoc-program-application-tips-179ceba24e6f?source=rss-a38c442737c0------2</link>
            <guid isPermaLink="false">https://medium.com/p/179ceba24e6f</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[tips]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[kaggle]]></category>
            <dc:creator><![CDATA[Catherine Liang]]></dc:creator>
            <pubDate>Tue, 08 Jun 2021 11:38:28 GMT</pubDate>
            <atom:updated>2021-06-08T11:41:39.320Z</atom:updated>
            <content:encoded><![CDATA[<p>Kaggle BIPOC Grant Program offers a great opportunity to grow your data science portfolio, with access to leading data scientists and a class of like-minded aspiring machine learning enthusiasts. (For more details on the learning from Kaggle BIPOC program, please read <a href="https://dailyimprove2012.medium.com/kaggle-bipoc-program-key-takeaways-d4f92f77acdb">my other article</a>). This article will highlight how to get accepted into the program.</p><ol><li>Show your passion for data science.</li></ol><p>Data science is a buzzword for a few years now. Are you passionate about it? If you are, what have you done for your passion. Do you perhaps have done a number of data science projects? Are you working in data science? Do you regularly read related research paper, blog, etc.? Do you blog about machine learning concepts?</p><figure><img alt="Courage The Cowardly Dog. Credit: Cartoon Network" src="https://cdn-images-1.medium.com/max/400/0*gsl75GKedUYr_HvA.png" /><figcaption>Courage The Cowardly Dog. Credit: Cartoon Network</figcaption></figure><p>2. What is your data science goal.</p><p>This is to help yourself to understand what you want to achieve in data science. It may be that you want to get into the field whether from your study or from your current career, or you want to move up the ladder in the machine learning world.</p><p>3. What difficulties you are facing to achieve your goal.</p><p>Now that you have worked out if you are truly passionate about this and the goal, consider what’s stopping you from achieving that objective. As a member of BIPOC, there are many difficulties (financial, socioeconomic, to name a few). Focus on the top one in your application and be specific. Is it because of the cost of having GPUs? Or that you don’t know who to talk to get some guidance?</p><p><strong>Conclusion</strong></p><p>I think having a clear response to each of these questions not only help you get a better chance in applications but also help you think through your current career path in data science. If you like this article, please give me a clap👏!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=179ceba24e6f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Kaggle BIPOC Program Key Takeaways]]></title>
            <link>https://dailyimprove2012.medium.com/kaggle-bipoc-program-key-takeaways-d4f92f77acdb?source=rss-a38c442737c0------2</link>
            <guid isPermaLink="false">https://medium.com/p/d4f92f77acdb</guid>
            <category><![CDATA[tips]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[kaggle]]></category>
            <dc:creator><![CDATA[Catherine Liang]]></dc:creator>
            <pubDate>Tue, 08 Jun 2021 11:37:03 GMT</pubDate>
            <atom:updated>2021-06-08T11:40:51.397Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="Photo by myself; SWAG is just the start (or rather the end)!" src="https://cdn-images-1.medium.com/max/1024/1*QmjACA8at5RhlVp-d970Uw.jpeg" /><figcaption>Photo by myself; SWAG is just the start (or rather the end)!</figcaption></figure><p>After a busy three months, I have finally made it through Kaggle BIPOC Grant Program! It has been an absolutely amazing journey and I would like to share my top learning. If you like to know how to apply, feel free to check out my <a href="https://dailyimprove2012.medium.com/kaggle-bipoc-program-application-tips-179ceba24e6f">other article</a>.</p><p><strong>Non-Technical</strong></p><ol><li>The power of team</li></ol><p>Don’t estimate the collective wisdom. Also don’t think you have nothing to bring to the table. We all have strengths and limitations.</p><p>2. Use the discussion board</p><p>Read the discussion board to check out other’s ideas and share yours if you want to. Kagglers are quite friendly; they won’t bite.</p><p><strong>Technical</strong></p><ol><li>Feature Engineering</li></ol><p>After this program, I would definitely focus on feature engineering more. Creating additional features has significantly boosted my model performance. But I have to admit, it is hard to come up with new features. That’s a top reason to team up or use the discussion board for new ideas.</p><p>2. Optuna</p><p>I learnt this from one of the mentors and tried in one of my deliverables. I spent 10 mins to adapt it to my use case and it returned 6 tuned hyperparameters in seconds! I was amazed as I had been waiting 2 hours for tuning 3 hyperparameters (all the time that I could have spent better).</p><p><strong>Conclusion</strong></p><p>This is by no means an exhaustive list of what I learnt during the program. But they are my top picks. If you like this article, please give me a clap 👏!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d4f92f77acdb" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Colab vs Google Cloud (free credit) in Kaggle Competition]]></title>
            <link>https://dailyimprove2012.medium.com/colab-vs-google-cloud-free-credit-in-kaggle-competition-d176b98940d0?source=rss-a38c442737c0------2</link>
            <guid isPermaLink="false">https://medium.com/p/d176b98940d0</guid>
            <category><![CDATA[gcp]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[kaggle]]></category>
            <category><![CDATA[colab]]></category>
            <dc:creator><![CDATA[Catherine Liang]]></dc:creator>
            <pubDate>Sun, 07 Mar 2021 01:57:42 GMT</pubDate>
            <atom:updated>2021-03-07T01:57:42.377Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mkLnnERqcdv4G-gyz9Tnhw.png" /><figcaption>Pick the right one for your project</figcaption></figure><p>If you’re trying to participate in a Kaggle Competition, or build machine learning models for your passion project, you’re likely to come across colab and/or Google Cloud Platform (GCP), among others. GCP offers a free introductory credit, sounds enticing? Which one is better for your passion project?</p><p>TLDR: Colab is better than GCP with free credit.</p><p>Here is the full story. I recently participated in a Kaggle Competition, that has a very large dataset (30+GB!). My team decided to give both a go and here’s what we found:</p><p>On Google Cloud free credit, you are quite limited in the types of instances you can use (so far, you can’t choose the ones with TPU). With the types of instances available, it can take a very long time to train and often run into memory error. In fact, I lost an entire instance (error 524) and had to create a new instance to start from scratch. In summary, with free credit, it can take a long time to train but no result is retrievable.</p><p>On the other hand, you can choose TPU in Colab and I trained the same and larger size model without issue. Granted, there are several OOM errors. But it is recoverable.</p><p>I would definitely go with Colab.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d176b98940d0" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>