<?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 Hassan Jalil on Medium]]></title>
        <description><![CDATA[Stories by Hassan Jalil on Medium]]></description>
        <link>https://medium.com/@hj576?source=rss-f7187ac23fcb------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*eHcywmf4F_0MgqCd.</url>
            <title>Stories by Hassan Jalil on Medium</title>
            <link>https://medium.com/@hj576?source=rss-f7187ac23fcb------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 21:46:49 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@hj576/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[Ramping up AWS Lambda with Docker]]></title>
            <link>https://aws.plainenglish.io/i-have-always-loved-leveraging-serverless-technologies-and-no-service-has-been-used-more-by-me-6c176e52261e?source=rss-f7187ac23fcb------2</link>
            <guid isPermaLink="false">https://medium.com/p/6c176e52261e</guid>
            <category><![CDATA[docker]]></category>
            <category><![CDATA[lambda]]></category>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[aws-lambda]]></category>
            <category><![CDATA[python]]></category>
            <dc:creator><![CDATA[Hassan Jalil]]></dc:creator>
            <pubDate>Wed, 24 May 2023 05:45:42 GMT</pubDate>
            <atom:updated>2023-05-29T01:12:27.436Z</atom:updated>
            <content:encoded><![CDATA[<h3>Ramping AWS Lambda Using Docker Images</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*s4-6nGQWn7Ajjgf-3WWq2w.png" /><figcaption>source: <a href="https://aws.amazon.com/blogs/compute/using-container-image-support-for-aws-lambda-with-aws-sam/">https://aws.amazon.com/blogs/compute/using-container-image-support-for-aws-lambda-with-aws-sam/</a></figcaption></figure><p>I have always loved leveraging serverless technologies, and no service has been used more by me than Lambda for a whole bunch of projects. Being able to run a script that can be triggered by a plethora of services and methods without having to worry about underlying infrastructure is amazing.</p><p>As someone who majorly works with Python, using lambda functions for scripts is something that always felt right, but it too had its limitations. Leaving the time and memory constraints aside (which also have increased a lot since its inception), the biggest problem we always faced with Lambda back in the day was dependencies. If you needed any third-party library, you had to bundle it with your code. You also had to create a zip with all the files of the library included. This would become even more difficult if the library you were using depended on C binaries. In these cases, one considered dropping Lambda for the solution altogether.</p><p>Frameworks like serverless frameworks made packaging libraries easier. They came with plugins that can be used to manage dependencies. You can specify your code and the framework was responsible for bundling all the Python libraries you mentioned in your requirement with it. This worked for most cases, but in cases where you may need additional C binaries, you were left scratching your head.</p><p>The solution is to create a docker image that was based on Amazon Linux. Inside the container, install the needed libraries and dependencies, copy over the library files and the needed C binary to your host machine, bundle them together and upload them to your lambda…… and your package is too big to be used by lambda, remove some c binaries which you think might not be needed and try again… and the library no longer works….</p><p>Yes, I went through this. For one project, I had to use a cryptography library to verify a SAML message on Lambda, and it took me HOURS and sleepless nights (s) to figure out which binaries needed to be bundled and which we could do without. For this purpose, my rule of thumb was not to run any script that had complex dependencies on lambda.</p><p>All that changed when in 2020, AWS allowed us to deploy lambda functions using docker images. All you have to do now is, create a Docker image that is based on Amazon’s Lambda base image, install all the needed libraries and dependencies, push the image to ECR and deploy a lambda function using the image.</p><p>The Docker image itself is pretty straightforward.</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/1ad97b743eec147724dd1d2f2dac5cf4/href">https://medium.com/media/1ad97b743eec147724dd1d2f2dac5cf4/href</a></iframe><p>FROM public.ecr.aws/lambda/python:3.9</p><p>We are using the base image for python lambda as provided by AWS</p><p>COPY requirements.txt /requirements.txt</p><p>We then copy over our requirements.txt</p><p>RUN python3 -m pip install -r /requirements.txt — target ${<em>LAMBDA_TASK_ROOT</em>}</p><p>We then install our dependencies</p><p>COPY ./code ${<em>LAMBDA_TASK_ROOT</em>}</p><p>We copy over our source code</p><p>CMD [“handler.lambda_handler”]</p><p>And finally, we tell which command needs to be run, when the lambda is executed. This points to the handler function where we have written code.</p><p>And using that simple Docker file, we can create an image that can then be used to run our Python code, along with its dependencies on Lambda.</p><p>We can add all the libraries we need in the requirements.txt. We can also install any third party C libraries or binaries we need to run our code in the Lambda environment. For example if you want to install psycopg2-binary, you can add the following command to install the libraries it needs.</p><p>RUN yum install -y postgresql-devel gcc*</p><p>I have created a simple sample Git repo that contains the most basic Dockerfile along with the code structure you need to create a docker image that can be used to deploy lambda functions.</p><p>Details on how to use this example code to deploy the imgae to lambda are given in the README.md</p><p><a href="https://github.com/hassanj-576/python-docker-example">https://github.com/hassanj-576/python-docker-example</a></p><p>If you want to create an even more complicated docker image, which is based on let us say ubuntu and not Amazon’s own Lambda image, you can even do that and use a library like awslambdaric (https://pypi.org/project/awslambdaric/). This lets you install Lambda’s runtime environment to any custom docker image.</p><p>Now, of course, this comes with its own drawbacks. The cold boot of a lambda running using a Docker image is slower than if you had uploaded a simple zip file. The docker image itself will be much bigger. So if you are writing a simple script with not many dependencies, simply create a zip file and upload it for your lambda code. On the other hand, if you are using something more complex, a script that depends on a number of Python libraries and C binaries, the best route is to use docker images. It is faster than having to manage all your dependencies manually and allows you to spend more time actually solving the problem at hand instead of pulling your hair out trying to make a lambda function work.</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=6c176e52261e" width="1" height="1" alt=""><hr><p><a href="https://aws.plainenglish.io/i-have-always-loved-leveraging-serverless-technologies-and-no-service-has-been-used-more-by-me-6c176e52261e">Ramping up AWS Lambda with Docker</a> was originally published in <a href="https://aws.plainenglish.io/">AWS 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[To ChatGPT or not to ChatGPT, that’s the Question!]]></title>
            <link>https://hj576.medium.com/to-chatgpt-or-not-to-chatgpt-thats-the-question-a6787cb3330d?source=rss-f7187ac23fcb------2</link>
            <guid isPermaLink="false">https://medium.com/p/a6787cb3330d</guid>
            <category><![CDATA[suggestions]]></category>
            <category><![CDATA[advice]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[computer-science]]></category>
            <dc:creator><![CDATA[Hassan Jalil]]></dc:creator>
            <pubDate>Wed, 26 Apr 2023 13:47:35 GMT</pubDate>
            <atom:updated>2023-04-26T13:47:35.139Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KwcsqiSw66QTLnWNfHzhsw.jpeg" /></figure><p>So here I am writing a blog after ages. I have always had ideas to write about but just too lazy to put pen to paper (fingers to keys?). Though for the topic at hand, I felt like I needed to write about this.</p><p>We have seen tremendous progress in the field of Artificial Intelligence specially Conversational AI. With ChatGPT seriously being a game changer and then with GPT 4 released, a lot of high profile people have stepped up talking about how the recent spike in AI advancement needs to be controlled. I understand why this rapid advancement in AI can be dangerous and how corporate greed can lead to problems down the road, but as things stand, AI is going to continue advancing at a very high pace.</p><p>As people from different professions fear for their jobs being replaced by AI, one has to understand it’s better to adapt than to fight. Lets take the example of chatGPT, as a programmer, I recently realised how useful ChatGPT can be for me (yes I am late to the party). For me, the approach whenever I hit a roadblock was the same as it has been since I started studying Software back in 2011. Google it and hope someone has the same question on stackoverflow or reddit. If not, go through the docs of the said library. This approach so far has paid dividends, but if the problem is a bit complex and not that common, it can take a lot of time to find the right solution.</p><p>I came across a similar problem recently when I was trying to use the rpy2 library that helps you run R using python. With very little academic experience of R, and that too from 5 years ago, finding the solution was proving to be difficult. Questions on stackoverflow did not seem to cover my problem. The docs made little sense as I was not really aware of R data types. After hours spent diving deeper into the never ending pit of google search results, I decided to give chatGPt a try. I asked my question in detail and waited as the cursor prompt blinked for a few seconds and then the AI started typing the answer.</p><p>The answer started off by giving a brief description of the problem at hand and its solution. It then followed it up with an example and actual code. The answer seemed adequate so I decided to copy and paste the provided code, modify it a little to suit my requirements and run the code. To my utter astonishment, it worked just as I wanted it to. At that time I realised that if opting to ask chatGpt for help was my first approach, I could have saved a couple of hours.</p><p>Now, ChatGPT isn’t perfect for another problem I faced. ChatGPT told me to pass an argument to a function in rpy2 library, which it does not accept. So it does make mistakes but you can tell it in reply that it did not work and it needs to give you the right answer and it will try again (though it may fail again). You can always rephrase your question and hope for better answers.</p><p>Since then, I have been using ChatGPT a lot more. If I hit a problem and Googling it does not return an answer on the first few links, I jump to chatGPt and let it have a swing. More often than not, our little AI friend gets the answer right.</p><p>What I found, chatGPT being very useful for is asking niche questions which cater for a particular situation. Questions for which you may have to dig into the library’s document to see how things are actually done to have a better understanding, chatGPt can save you hours of research by giving you a very direct answer. Now granted, you should take the answers with a grain of salt since there is a possibility the answer may be wrong but in most cases, it proves to be quite accurate.</p><p>The basic idea that I am trying to share is; AI advancement at an insane pace is most probably going to continue. The fear is somewhat justified but shunning a technology out of fear of what it may impact down the road is not the right solution. One needs to adapt and use these technologies for their own benefit. If chatGPT can help me reduce my search for a solution from 30 down to 5 minutes, that is an additional 25 minutes I have in the day to be more productive. Use these technologies for your own good, utilise them to increase your efficiency and get more stuff done. We need to adapt and adopt the latest tech and utilise it for our own benefit.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a6787cb3330d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Creating a T20 Prediction Twitter Bot 1/2]]></title>
            <link>https://medium.com/analytics-vidhya/creating-a-t20-prediction-twitter-bot-1-2-591a992e8694?source=rss-f7187ac23fcb------2</link>
            <guid isPermaLink="false">https://medium.com/p/591a992e8694</guid>
            <category><![CDATA[bots]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[cricket]]></category>
            <dc:creator><![CDATA[Hassan Jalil]]></dc:creator>
            <pubDate>Wed, 03 Mar 2021 08:54:30 GMT</pubDate>
            <atom:updated>2021-03-04T16:50:35.454Z</atom:updated>
            <content:encoded><![CDATA[<h4>Using Historic T20 Match data to create a win prediction bot that posts predictions on Twitter</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*--qjcndUmj5ocGc1dWqp4Q.png" /><figcaption>Photo by <a href="https://unsplash.com/@ajayparthasarathy">Ajay Parthasarathy</a></figcaption></figure><p>This is the 1st part of a 2 part series that discusses how I made a Twitter Bot that makes predictions on who is going to win a T20 match. In the first part, we will discuss the machine learning aspect of the bot, from getting the data, transforming it, and training the model. In part two we will discuss the Twitter bot, how it gets the match data, and tweets the prediction in real-time.</p><h3>Cricket Fever</h3><p>My cricket fever is still at an all-time high, there is something about Pakistan Super League that brings out the Crazy Cricket fan in me. After scrapping ball by ball data of all previous PSL seasons (You can read that <a href="https://medium.com/swlh/pakistan-super-league-data-set-6b04f15cb86c">here</a>) and then running some basic data analysis and creating visualization on that data set ( you can read that <a href="https://hj576.medium.com/pakistan-super-league-data-analysis-c366e7c198a4">here</a>), I wanted to do more with the dataset. Then it hit me, why not train a model that would predict the likelihood of a team winning the match.</p><h3>Motivation</h3><p>Those who watch cricket would know that most modern-day broadcasts have a win predictor that makes a prediction on what percentage chance a certain team has of winning. Looking at such a predictor on the broadcast of PSL made me think of making my own</p><h3>Disclaimer</h3><p>The predictor I have made is VERY BASIC. It does not account for a lot of factors that make a huge impact on the match, such as the venue, pitch, weather, form of players, and teams. All it looks at is how many runs are needed, how many balls are in hand, and how many wickets have fallen. Owing to my busy schedule, making a more detailed predictor is not really feasible and I was more interested in the whole process of making one than the actual results.</p><h3>Data Set</h3><p>What I wanted to teach my model is, given a certain scenario, what are the chances that the team chasing will win. For this, I needed ball by ball data for matches, such that I would know at any given ball of the second inning, what the score was, and whether the team chasing won or not.</p><p>I started with the PSL dataset that I scrapped myself ( <a href="https://medium.com/swlh/pakistan-super-league-data-set-6b04f15cb86c">link</a> to the post). This contained ball by ball details for both the innings for all matches of PSL. I trained the first model on this dataset, but since PSL is relatively new and each season has a maximum of 34 matches, this dataset was too small(33962 rows total).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*oBPR18-hNjGOStE5mwbmUA.png" /></figure><p>Next, I decided to look for another dataset. I came across <a href="https://www.kaggle.com/patrickb1912/ipl-complete-dataset-20082020">this</a> dataset on Kaggle which contained ball by ball detail for each match of the Indian Premier League. IPL has been running for over 10 years, has far more teams and far more matches per season than PSL. This gave me a somewhat adequate dataset (179080 rows).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rQisoKA-AD4-AuwtaNwcwg.png" /></figure><h3>Transforming The Data Set</h3><p>For solving a machine learning problem, the first step is to gather a dataset, the next is to clean it and transform it into a form that you can feed to your machine learning algorithm. Now with the data downloaded, I needed to transform it.</p><p>The predictor will only make predictions in the second inning of the match when the target has been set and the chasing team has started batting. My dataset contains ball by ball data for both innings. It also does not contain the target. Moreover, the IPL dataset did not contain values for who won the match in their ball by ball dataset.</p><p>The first step was to calculate the first inning total for each match, this would give us the target that the second team had to chase. With pandas, we can easily achieve this.</p><p>We first generate a data-frame that consists only of first innings data. We then group by the match id and aggregate the run column by applying a sum</p><pre>first_inning_total = first_inning.groupby(&quot;match_id&quot;).agg(<br>    target = (&quot;total_runs&quot;,&quot;sum&quot;)<br>)</pre><p>This will sum the first-inning runs for each match and give us the total runs scored in the first inning, which is the target the second team has to chase.</p><p>The IPL dataset came with two tables, ball by ball details table, and a match summary table. Who won the match was given in the match summary table. So by joining both tables on the match_id we were able to get the name of the winner of each match along with its ball by ball detail.</p><p>Now we have, the target, ball by ball data for the second inning, and who won the match. We simply apply some more simple dataset-specific transformation to calculate things such as the number of fallen wickets, run scored, target, overs, and balls left e.t.c and we have our dataset ready. (All of the code is on GitHub and the link is shared at the end of this article)</p><p>By combining both the IPL and PSL dataset, keeping second inning data only we end up with 102560 rows that we can train our model on.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/632/1*n_kpM67HKvKPNvDvcgGisg.png" /><figcaption>Final Form of the Dataset</figcaption></figure><h3>Features</h3><p>As you can see in the image above, we are just using wickets balls_left and runs_left as our features for the model. In a proper model, we would have a lot more features. Who is on the crease, who is bowling, type of pitch, weather, the form of both the teams and so on? The possibilities of the available features are quite large, this is a very simple bot based on the simple feature.</p><p>Initially, the first model I trained also used more features. The features set included runs_scored, overs, ball, and target . The performance of the model I trained on these features wasn&#39;t that great in real life. Even though the accuracy on the test set was over 80%, in real-life performance the model did not make good predictions. The model was being swayed a lot by wickets, even with 5 runs required in the last over with only 5 wickets down, the model was predicting the bowling side would win. If you change the wickets from 5 to 4, the model would predict the batting side would win. It seemed like the model had learned that the higher the wicket number, the lower the chances of winning, while it seemed to not get the relationship between target and score. So I decided to simplify the features. Instead of target and score as separate features, I used runs left (by subtracting the score from target) instead of using over and balls, I used the feature balls_left.</p><p>With these new features, the model seemed to perform much better in real-life conditions. The model was able to understand that the lower the run left, the more the chance of the team batting winning. On the other hand if the batting team have more balls left, their probability of winning is higher. While fallen wickets result in lowering your win percentage. The model was able to understand all these features and the relationship between them.</p><p>So by using smaller and simpler features, our model was able to perform model.</p><h3>Training the Model</h3><p>The problem can be simplified to simple binary classification. All we need to know is whether the team chasing won or not. So we can simply predict 0 or 1, 0 being the team lost and 1 being the team won.</p><p>The first step is to divide our dataset into X and y, where X is the feature and y is the result.</p><pre>y = full_df.iloc[:,3]<br>X = full_df.iloc[:,:3]<br></pre><p>Next is to split the dataset into test and train sets.</p><pre>X_train, X_test, y_train, y_test = train_test_split(<br>    X,<br>    y,<br>    test_size=0.2,<br>    random_state=42)</pre><p>For the algorithm, since this is a simple binary classification, I just decided to use a Random Forest Classifier.<br>So the training can be run with the following code</p><pre>RF = RandomForestClassifier(n_estimators=10000, max_depth=6, random_state=0)<br>RF.fit(X_train, y_train)</pre><p>This will run training on the dataset. The training is quite quick and took a little over 3 minutes on my laptop.</p><p>Once the model has been trained, we can evaluate it on our test dataset.</p><pre>y_pred_test = RF.predict(X_test)<br>accuracy_score(y_test, y_pred_test)</pre><p>We got a score of 0.77 which for such a basic model seems good enough.</p><h3>Evaluation</h3><p>To get a prediction, all we have to do is pass it a data frame in which each row contains the features we have trained our model on and call the predict method on the model file.</p><pre>current= {<br>    &quot;wickets&quot;:1,<br>    &quot;balls_left&quot; :105,<br>    &quot;runs_left&quot;:137<br>}<br>current_df = pd.DataFrame(current,index=[0])<br>RF.predict(current_df)</pre><p>The predict method returns the label the model has predicted(0 or 1 in our case), we though want to predict percentage chances of a win, so we can use the `predict_proba` method instead, this returns the probability of either of the labels being the outcome.</p><pre>RF.predict_proba(current_df)<br>array([[0.41718799, 0.58281201]</pre><p>This basically means there is a 41% chance the team batting is going to lose (or the first team is going to win) and a 58% chance that it will win. Thus based on these results we can easily make our percentage prediction.</p><h3>Real Life Result</h3><p>My predictor (along with the Twitter bot) has been running for quite some days now and has been making its prediction on all live T20 matches. And honestly, I am quite surprised at how good the predictions have been. Now predicting a cricket match, especially a T20 match is not easy, even with a big data set, a complex model, and all kinds of features, you cannot take into account when Afridi is going to go berserk and win you a lost match single-handedly. <br>What I was more interested in, was how my predictor reacted to changes in matches. Based on the current situation of the match can it make the right prediction, even though the prediction may not hold till the end of the match due to some match-changing performance.</p><p>Based on these metrics, my predictor actually performed quite well. For a given scenario it made predictions that made sense. If the chasing team was scoring well, had wickets in hand it would make the right prediction of the chasing team having a higher probability of winning.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/605/1*rExuRbRGjUSR-OzeHCI-MQ.png" /></figure><p>On the other hand, if the target was high, and the chasing team had lost early wickets with not many scores on the board, it would rightly predict the bowling team had a higher chance of winning</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/606/1*naWKzU_nA0DYmnqIRKHBHA.png" /></figure><p>It also catered well to changing scenarios. In case the chasing team needs 25 runs in 11 balls, has already lost 5 wickets, it predicts the bowling team has a slightly higher chance of winning</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/536/1*qYHdyFVSxc3MFV04GUApuw.png" /></figure><p>But if Wahab Riaz out of nowhere decides to hit Steyn for two massive sixes bringing the target in arms reach, it quickly adjusts its prediction as well.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/599/1*R2VYgTwv8qV7epSJuWOXcQ.png" /></figure><p>Across the multiple matches the predictor has ran on, the results have been outstanding. The predictor adjusts well for changing scenarios and the predictions are based on the current situation. With a little more tweaking and using maybe a more complex algorithm, we may be able to achieve even better predictions</p><h3>Next Part</h3><p>In the next part of this series we will discuss how I created a Twitter bot, that waits for a match to start, gets its score, and then makes a prediction and tweets it. Once the post is up I will link it here</p><p>Thank you for reading the article, if you have any questions feel free to reach out to me on my LinkedIn or Twitter</p><h3>Links</h3><p>PSL Data Set: <a href="https://www.kaggle.com/hassanj576/pakistan-super-leaguepsl-ball-by-ball-20162020">https://www.kaggle.com/hassanj576/pakistan-super-leaguepsl-ball-by-ball-20162020</a></p><p>IPL Data Set: <a href="https://www.kaggle.com/patrickb1912/ipl-complete-dataset-20082020">https://www.kaggle.com/patrickb1912/ipl-complete-dataset-20082020</a></p><p>Code for Transforming and Training: <a href="https://github.com/hassanj-576/t20_predictor">https://github.com/hassanj-576/t20_predictor</a></p><p>My LinkedIn: <a href="https://www.linkedin.com/in/shjalil/">https://www.linkedin.com/in/shjalil/</a></p><p>My Twitter: @hj576</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=591a992e8694" width="1" height="1" alt=""><hr><p><a href="https://medium.com/analytics-vidhya/creating-a-t20-prediction-twitter-bot-1-2-591a992e8694">Creating a T20 Prediction Twitter Bot 1/2</a> was originally published in <a href="https://medium.com/analytics-vidhya">Analytics Vidhya</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Pakistan Super League Data Analysis]]></title>
            <link>https://hj576.medium.com/pakistan-super-league-data-analysis-c366e7c198a4?source=rss-f7187ac23fcb------2</link>
            <guid isPermaLink="false">https://medium.com/p/c366e7c198a4</guid>
            <category><![CDATA[cricket]]></category>
            <category><![CDATA[data]]></category>
            <category><![CDATA[data-visualization]]></category>
            <category><![CDATA[data-analysis]]></category>
            <dc:creator><![CDATA[Hassan Jalil]]></dc:creator>
            <pubDate>Wed, 24 Feb 2021 11:00:39 GMT</pubDate>
            <atom:updated>2021-02-24T11:00:39.480Z</atom:updated>
            <content:encoded><![CDATA[<p>An interesting data analysis of PSL ball by ball data from 2016–2020</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xlS8XNC7guSTHLmtceTjow.jpeg" /><figcaption>PSL Opening Cermoney</figcaption></figure><p>Continuing from where I left off in my <a href="https://hj576.medium.com/pakistan-super-league-data-set-6b04f15cb86c">previous post</a> of extracting PSL Data, I decided to crunch the numbers and see what interesting analysis i can extract from the data. <br>Given the limited dataset, and the limited free time I have, I was able to run some basic analysis and create some pretty charts. So here it is</p><h3>Boundaries Per Season</h3><p>As we know, the first edition of PSL took place entirely in UAE while the Final of the next Edition took place in Pakistan, this saw a trend of moving more and more matches to Pakistan with the 2020 edition being completely hosted in Pakistan (yaaaay)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*_a32uIAoxNojZVoVoI9rBg.jpeg" /><figcaption>Captains Posing with the Trophy in Pakistan before the start of the tournament</figcaption></figure><p>From a spectator point of view, the pitches in UAE seemed more slow and we saw lower scoring total, so it was interesting to see how many boundaries were hit in each season.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/298/1*0QMmjRbrKItaQctOhoS6wA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/521/1*eJwejLp_BQuJxGOPQJbl9w.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/521/1*jqy0TKI6jYaX_4GRUhCXbg.jpeg" /></figure><p>As you can see, the first two seasons had far less boundaries than the next three, but this is also owed to the fact that first two seasons had 5 teams while Multan Sultan was introduced in the third season. This resulted in first two seasons having a total of 24 matches, while the next three had 34 matches each. <br>Now If we account for the number of matches, and calculate number of fours and sixes per match we get the following figure</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/838/1*o2di6u-FDY2iP21KJv4eGw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/1*KcykixHuWZnvTkZjLapL8w.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/1*Jtg_XPh00kNLj1urIQGhYg.jpeg" /></figure><p>It is surprising to see that the number has gone up, specially for fours, but it has not gone up a subsequent amount. I was expecting due to slower paced pitches of UAE, we would see far less boundries.</p><h3>Wickets</h3><p>PSL has been famous for its bowling standards. Mostly we have seen low to medium scoring games with very few games going close to the 200 marks. It is interesting to see what kind of wickets were the most common in PSL</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/446/1*Vi8HV2wwqkn8FLaWCEZ6xw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BTihKHTeI_mfv0Uq8PEZYw.png" /></figure><p>As expected, in a game like T20, the most amount of wickets fell via catches.</p><h3>Team Runs</h3><p>Another thing, I wanted to look into was how many runs were scored by each team.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*27WJGBreG63-ap3bpKOY2w.png" /></figure><p>It can be see Peshawar Zalmi scored the most amount of runs (8632) but this can be owed to the fact they have played the most matches thanks to them qualifying for the playoff every year. Also Multan did not play the first two seasons, hence their bar is the lowest. If we account for runs scored per match, we get a different picture.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DzKNpxgZ-esMeYrns4_dpw.png" /></figure><p>This Graph shows, though Peshawar is still slightly ahead, all the teams are much closer and have almost scored the same amount of runs per match. I believe this is due to the fact that T20 games are much closer and usually the win margins aren&#39;t that big, also team playing second would never really score a lot more than the team playing in the first inning even if they win easily.</p><h3>Percentage Runs Scored in Boundaries</h3><p>The next thing I wanted to analyse is something I believe is a big problem faced by Pakistani Cricket, our players do not rotate the strike and mostly score in boundaries. They play a lot of dot balls and then hit a boundary to keep their strike rate high. So I wanted to check what percentage of runs were scored by each team from boundaries only</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*XstXIZtrYXwCmUffoUvJ6g.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6Ujz2xbFqtF871y1s08F2g.png" /></figure><p>As you can see, this number is quite high. All teams have scored over 50% of their runs via boundaries. with Islamabad United scoring almost 59% of the runs.</p><h3>Dot Balls</h3><p>Now in a game like T20, it is expected a lot of runs will be scored via boundaries and batsman will hope to go for the maximum, but I believe, our batsman rely solely on boundaries and do not rotate the strike and play a lot of dot balls. So here is the Analysis of what percentage of balls did not go for any runs for each team.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/984/1*N1OFB6bzjXXMPJCq-i0EBg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*eMtAJRMqyV3yumLa3900hw.png" /></figure><p>As you can see, Islamabad who made the most number of runs via boundaries, also have almost the highest percentage of dot balls. Every team is playing at least 35% dot balls, that is at least 7 overs in a 20 over match with no runs scored. And I believe this is one of the biggest problems plaguing Pakistan cricket. In modern day cricket with average runs per match increasing every year, Pakistan team needs to make every ball count and rely on taking singles and doubles beside just hitting boundaries.</p><h3>Sixes Per Team</h3><p>Finally, just for the fun of it, lets look at the number of sixes each team has hit. Again since each team has played different number of matches, instead of calculating total sixes, we will calculate sixes hit per match on average</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*nHGNDD1ACl6P-LgnJp7A7Q.png" /></figure><p>Islamabad, who score the most percentage of runs in boundaries have hit the most number of sixes per match, followed by Peshawar and Lahore. I guess when you have had Luke Ronchi in your squad, the number sixes are bound to go up (he hit 50 sixes across his career in PSL)</p><h3>Ending Notes</h3><p>This was a little quick analysis of the data I gathered for the PSL from 2016 to 2020. I am sure there is a lot more information one can extract from this data and do a much better analysis.</p><p>If Anyone wants to take a swing at it, the data can be found on kaggle : <a href="https://www.kaggle.com/hassanj576/pakistan-super-leaguepsl-ball-by-ball-20162020">https://www.kaggle.com/hassanj576/pakistan-super-leaguepsl-ball-by-ball-20162020</a></p><p>If you have any questions regarding the data, feel free to contact me on <br><strong>LinkedIn : </strong><a href="https://www.linkedin.com/in/shjalil/">https://www.linkedin.com/in/shjalil/</a></p><p><strong>Twitter : @</strong><a href="https://twitter.com/hj576"><strong>hj576</strong></a></p><p>The code for this analysis, and scrapping the PSL data can be found on the git repo : <a href="https://github.com/hassanj-576/psl_cricinfo_scrapper">https://github.com/hassanj-576/psl_cricinfo_scrapper</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c366e7c198a4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Pakistan Super League Data Set]]></title>
            <link>https://medium.com/swlh/pakistan-super-league-data-set-6b04f15cb86c?source=rss-f7187ac23fcb------2</link>
            <guid isPermaLink="false">https://medium.com/p/6b04f15cb86c</guid>
            <category><![CDATA[data]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[kaggle]]></category>
            <category><![CDATA[cricket]]></category>
            <dc:creator><![CDATA[Hassan Jalil]]></dc:creator>
            <pubDate>Tue, 23 Feb 2021 11:59:55 GMT</pubDate>
            <atom:updated>2021-02-28T09:26:11.310Z</atom:updated>
            <content:encoded><![CDATA[<p>Ball by Ball Data of Pakistan Super League from 2016 to 2020</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/800/1*Y-5RPgkX1Kono-lmjNRuZA.jpeg" /><figcaption>Pakistan Super League</figcaption></figure><p>Those of you, who know me, know how big of a Cricket fan I am. I grew up watching this sport, and even though I am not a purest who watches all the matches (I just watch the ones involving Pakistan), I still have undying love for this game.</p><p>Pakistan Super League (Pakistan’s version of T20 Franchise League) just kicked off its sixth season and I couldn’t be more excited! A month of amazing cricket featuring our favourite crickets from around the globe.</p><p>I was going through LinkedIn one day and saw a <a href="https://www.linkedin.com/posts/sheikh-haris-khalid-951403b9_psl-python-pakistan-activity-6768044284051636224-LVhu">post</a> from one of my contacts regarding a PSL dataset that he had generated. He shared a sample of it and said he would publish it soon. The Dataset seemed quite comprehensive and gave me ideas about different visualisation that can be generated from the data set along with some predictive models one can train.</p><p>I waited a day or so for him to publish the dataset, and when he didn&#39;t, I decided to start working on my own. So here I present, the Pakistan Super League data for all the matches from the first edition back in 2016 to the the latest one in 2020/2021</p><p>The Data has been scrapped from <a href="https://www.espncricinfo.com/">ESPN Cricinfo</a> website and contains ball to ball details for every match. The data is stored in a CSV and looks as following</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/997/1*6bGz5MhIUKS4jg8opG6jhQ.png" /></figure><p>We have 16 columns, most of them are pretty self explanatory, details of some of them are giving below</p><ol><li>is_four : Is set to true in case a four was hit</li><li>is_six : Is set true in case a six was hit</li><li>wicket : a numerical categorical representation of the type of wicket (cricinfo’s internal categories)</li><li>wicket_text : a textual categorical representation of the type of wicket (cricinfo’s internal categories)</li><li>result : Who won, or if the match was abandoned or tied</li></ol><p>There are a total of <strong>33962</strong> rows covering a total <strong>126 matches </strong>ball by ball.</p><p>As far as the script to scrape this data is concerned, I have published it on my GitHub and it can be found <a href="https://github.com/hassanj-576/psl_cricinfo_scrapper">here</a>. The script is quite simple. After inspecting Cricinfo’s website, I extracted the internal API call they make to load commentary dynamically. This API then returns ball by ball result for 2 overs. So getting all the data was as simple as recursively calling this API for each over and extracting the data. <br>The Api call takes the series_id, match_id and inning_numberas inputs along with the over from which we want the result.</p><p>So the API URL looks like this</p><p><a href="https://hs-consumer-api.espncricinfo.com/v1/pages/match/comments?seriesId={series_id}&amp;matchId={match_id}&amp;inningNumber={inning_number}&amp;commentType=ALL&amp;fromInningOver={over_from}">https://hs-consumer-api.espncricinfo.com/v1/pages/match/comments?seriesId=<strong>{series_id}</strong>&amp;matchId=<strong>{match_id}</strong>&amp;inningNumber=<strong>{inning_number}</strong>&amp;commentType=ALL&amp;fromInningOver=<strong>{over_from}</strong></a></p><p>If you know the series_id and match_id you can easily extract information for each ball.</p><p>So, with simple web scrapping and extracting the series and match id for each matches in a PSL series, I was able to extract the needed data.</p><h3>Future Work</h3><p>The data scrapped at the moment is quite basic. I patched this up in a single night. I would love to extract more data. At the moment one big thing missing is player information. I would love to add who was the bowler and who was on strike so we can have player wise data as well.</p><p>Besides improving the scrapped data, I plan to run some visualisation on this dataset to get more insight regarding the tournament. I also plan to work on a match predictor (similar to what we see on TV though a lot more simple), though I will definitely need a lot more data for it.</p><p>I will keep publishing new articles as I make more progress, so stay tuned folks !</p><p>You can reached out to me on my LinkedIn : <a href="https://www.linkedin.com/in/shjalil/">https://www.linkedin.com/in/shjalil/</a></p><p>or on twitter @hj576</p><p><strong>Link to Data set </strong>: <a href="https://www.kaggle.com/hassanj576/pakistan-super-leaguepsl-ball-by-ball-20162020">https://www.kaggle.com/hassanj576/pakistan-super-leaguepsl-ball-by-ball-20162020</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6b04f15cb86c" width="1" height="1" alt=""><hr><p><a href="https://medium.com/swlh/pakistan-super-league-data-set-6b04f15cb86c">Pakistan Super League Data Set</a> was originally published in <a href="https://medium.com/swlh">The Startup</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Quick Weekly Tech News : 1st Week of July]]></title>
            <link>https://hj576.medium.com/quick-weekly-tech-news-1st-week-of-july-32a53bfc9c9e?source=rss-f7187ac23fcb------2</link>
            <guid isPermaLink="false">https://medium.com/p/32a53bfc9c9e</guid>
            <category><![CDATA[news]]></category>
            <category><![CDATA[tech]]></category>
            <category><![CDATA[microsoft]]></category>
            <category><![CDATA[tiktok-app]]></category>
            <dc:creator><![CDATA[Hassan Jalil]]></dc:creator>
            <pubDate>Mon, 06 Jul 2020 12:26:44 GMT</pubDate>
            <atom:updated>2020-07-06T12:26:44.849Z</atom:updated>
            <content:encoded><![CDATA[<h3>Quick Weekly Tech News : 1st Week of July</h3><p>So I have started a Youtube channel where every Sunday I post a quick 2 minute look back at the weeks Tech News. I have decided to share the news here as well in form of a blog. <br>If you are interested in the video, its given below</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F5No_-Gq7giY%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D5No_-Gq7giY&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F5No_-Gq7giY%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/00b682f336a787ba078b358243fc7c0f/href">https://medium.com/media/00b682f336a787ba078b358243fc7c0f/href</a></iframe><p>So moving onto the tech news</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dwNMmwpB0rF2DJGUWoMcmg.png" /></figure><p>Microsoft has announced that it will be closing down all their physical stores in the world except 4 cities. The company will move to digital retail. Microsoft will not be laying off any employees as a result of this closure. Originally the plan was to close the physical store next year but due to corona the closure was announced earlier than initially planned.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NQRNWglZnFx-kTn8VB6vVg.png" /></figure><p>Microsoft also announced an all new Windows 10 start menu. The start menu is getting an overhaul as it looks much cleaner now. The new update is made available to windows 10 testers. Microsoft is introducing a more streamlined design that removes the solid color backplates behind the tiles and applies a uniform background.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2bsphnoFM2qmaf39RSm3tg.png" /></figure><p>Intel’s 11th Gen Rocket Lake processor was spotted on GeekBench, The processor in question has 8 cores and 16 threads. The processor has a base clock of 3.2 Ghz that Boost ups to 4.3 ghz. The chipset is using Intels XE Graphics for its integrated GPU. The initial benchmark scores aren’t too promising. Things are expected to improve as it is still in development.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*buJMr8tt9Df57DScPMwOnw.png" /></figure><p>As a result of stand off between China and India in the disputed boarder lines and the ensuing tension between the two countries, India has decided to ban 59 chinese apps including TikTok. The move comes in an attempt as India tries to hit China economically. India has over 120 Million Active Tik Tok Users and the app was downloaded in India over 450 Million times. Banning TikTok and other app will surely hurt China, though on the other hand it has also effected a lot of Indians that were using TikTok as their source of income.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iBevIMjYROaZwR5y2M-xWQ.png" /></figure><p>iOS new update tells users when an app is copying stuff off the clipboard. A user found out that TikTok was constantly copying off stuff from users’ clipboards. This news gained a lot of popularity and people started raising questions regarding TikTok violating their privacy. It was later found out its not just TikTok that does this but many other apps. So far 53 apps have been found including New York Times app, Fruit Ninja and Fox news.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=32a53bfc9c9e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Lack of Research in Pakistan’s Academic Institutes]]></title>
            <link>https://hj576.medium.com/lack-of-research-in-pakistani-academic-institutes-aa4a71fdffa3?source=rss-f7187ac23fcb------2</link>
            <guid isPermaLink="false">https://medium.com/p/aa4a71fdffa3</guid>
            <category><![CDATA[research]]></category>
            <category><![CDATA[university]]></category>
            <category><![CDATA[education]]></category>
            <category><![CDATA[pakistan]]></category>
            <dc:creator><![CDATA[Hassan Jalil]]></dc:creator>
            <pubDate>Tue, 08 May 2018 08:48:35 GMT</pubDate>
            <atom:updated>2018-05-08T08:49:00.038Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OFTCGYe8fKzFFjMOOCjHaw.jpeg" /></figure><p>I have been meaning to write this blog for quite sometime now, then again I have been meaning to write a lot of blogs as well. What finally pushed me to actually sit down and start mashing my fingers against the keyboard is two recent post.</p><p>First, this post about how IIT Delhi will now allow its PhD student to start a startup instead of submitting a Thesis.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/489/0*YwjEywDp5M5xw14J." /></figure><p>And this email from our University informing us about a session on how to monetize your startup</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*lBSBzqDt1LSN55Fk." /></figure><p>So, what am I complaining about this time? These two seem quite reasonable posts, shouldn’t we be promoting entrepreneurship as its’ the backbone of the economy. Yes it is, there is no doubt about it. But at the same time, research is the backbone of an Academic Institute. We see people complaining a lot about how Pakistani institutes are not higher up in the ranking, how there is hardly any research being carried out by our top institute, student do not seem to be focused on researching and there are hardly any research papers getting published by undergrad students.</p><p>The problem is, we are never really told about research. At least for me, during my Bachelors at NUST, we get a lot of preaching about entrepreneurship but nothing about research. In our Introduction to ICT course, we were introduced to this new idea of entrepreneurship and it was then reinforced throughout our bachelors. We were told about Steve Jobs and Bill Gates. Alumni came and gave us talks about how their startups are doing. Our heroes were always these amazing people whose startup turned into billion dollar companies. We even had an Entrepreneurship course in our 7th semester where we were supposed to run a small business or have an entrepreneurial venture to generate positive cash flow. Hence almost everyone I knew, at one point or another, tried to create their own “startup”, 99% of them didn’t even survive a year.</p><p>Not once were we told about research. We were never told to idolize people who have contributed to the field of research. We only knew Dijkstra as someone who made a shortest path algorithm, and not as some one who has over 1200 publications with 56463 citations (<a href="https://scholar.google.com.br/citations?user=g875mLgAAAAJ&amp;hl=en">https://scholar.google.com.br/citations?user=g875mLgAAAAJ&amp;hl=en</a>). Research was something reserved for the nerds and only for PhD students.</p><p>Throughout my bachelors, I read a total of 3–4 research papers, even those I did not really understand. Now most student do end up consulting more research papers specially during their Final Year Project, but everyone I have talked to had difficulty reading and comprehending research papers. We did not have a single course that focused purely on research and its aspects. A course that could teach us how to read a research paper and more importantly how to write one and how important research is to pursuit higher education. We need to be taught about the impact research makes in the field of science, and how it can help gain recognition not only for you, but your institute and country as well.</p><p>During my masters at University of Amsterdam, I had to read tons of research papers. I found it quite challenging in the start while my European colleagues had no problem picking up a research paper and going through it. In every course we were told about what’s the latest research being carried out in that particular discipline and what are the professors’ own current research topics. We were told about notable people who carried out research in the said field. We even had a course where we had to write a review on the research that has been carried out on the topic that our supervisor gave us. During my internship at CWI for my master thesis, I actually enjoyed doing research. I was surrounded by these super smart PhD students and just sitting down and discussing a research topic or paper was actually quite fun (yes i sound like a nerd right now but trust me it isn’t that bad). The reason was, we had this environment where research was emphasized on. We discussed ideas, had a “secret reading club” where we would discuss a research paper every month. We were given the perfect platform to research and have fun while doing it (not to mention we were getting paid and had an awesome office with even better hardware)</p><p>I wouldn’t deny that doing research is not an easy task. When you first pick up a research paper, reading it can be quite a daunting task. The language used, the structure and everything can seem something that a normal person can not comprehend. Writing one and getting it approved for publication is even more difficult. Also, research is not for everyone nor does everyone aspire to pursuit higher education. Same can apply to entrepreneurship. Yet the Pakistani universities seems to focus a lot on the latter and not at all on the former.</p><p>As I said earlier, research is of utmost importance for an academic institute. I agree that what a person learns should be practically applicable, and startups can be quite beneficial,and promoting entrepreneurship at institutes is quite important, but this should not come at the expense of research. For a university, research is the main goal (hence it has the highest weightage in calculating QS University rankings). Yet, our institutes seem to focus more on pushing people towards startups, instead of encouraging them to do research. As posted above, my institute seems to be holding a seminar on how to monetize your FYP, though I highly doubt they have held one on how to publish research papers (I haven’t came across one). We get industrial guests and entrepreneurs as guest speakers all the time, while no prominent researcher is invited to discuss their research topic.</p><p>Our institutes need to reevaluate what’s important. We need to glorify research as we have been glorifying entrepreneurship in the past decade or so. We need to introduce courses that aim at teaching students about the latest research topics, and how to dive into the world of research. Students need to be encouraged to read research papers, having a course where you write a position paper on a couple of research papers in your 3rd year could really help students get started on this journey. Invite researchers to tell students about their work, hold workshops on new research topics, build better labs, offer internships, fund projects, most importantly, the professors themselves need to take part in research and encourage bachelor students to join them.</p><p>Only through these drastic steps, and a complete overhaul on what our institutes think is important, can we hope to have our universities higher up in International rankings and gain international recognition.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=aa4a71fdffa3" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>