Last year we took our annual data science survey to the next level by turning over the results to YOU through an open-ended Kernel competition.
We were overwhelmed by the response and quality of kernels submitted. Not only are Kagglers amazing data scientists, but they’re incredible storytellers as well!
Mhamed Jabri was one of those skillful enough to take our data and shape it into something meaningful— not just for Kaggle, but for the data science community at large. We hope you enjoy getting to know him as much as we did.
Congrats, Mhamed, on your win!
To take a look at Mhamed's winning Kernel, visit: AfricAI
What was your background prior to entering this challenge ?
I’m currently pursuing an MS in Applied Mathematics at Ecole Centrale Lyon. My first formal background in machine learning came through several MOOCs and internships that I had during my gap year. I became active on Kaggle last year, winning multiple Kernel Awards and finishing top 1% in the Data Science Bowl 2018 competition.
What made you decide to enter?
Multiple reasons actually. A couple of months ago, I had the chance to attend Deep Learning Indaba in South Africa. The Indaba is a week-long conference that aims to strengthen machine learning in Africa through state-of-the-art teaching and networking in a very inclusive and diverse environment. My experience there was fantastic and made me realize how good African researchers really are. Since then, I was looking for a way to not only share that experience but also talk about AI in Africa in general and showcase the wonderful things that have been going on / that are happening right now in that continent for people who might be overlooking it. So the moment Kaggle sent us the first email about the survey to be conducted, I decided that when the results of the survey will be public I'd use them and build a storytelling notebook, which resulted in « AfricAI ».
What was your most important insight into the data?
Hmmm, I’d probably say that the most important insight is the one explaining how important MOOCs and online content are for students in Africa, especially in ML. Not only does it show from their answers in the surveys (as shown in my notebook, most African students think that online resources are better than what they’re given in class) but also from multiple success stories that I always encounter in Twitter, such as the one that I’ve taken from Jeremy Howard’s feed.
The most popular way to start learning Data Science / Machine Learning is Online Courses.
Were you surprised by any of your insights?
So there’s one insight but I think that for me, it was more of a disappointment rather than a surprise: The fact that less than 5% of respondents come from Africa. I definitely hope that the Kaggle community will have more Africans in its ranks in the years to come. It would also be amazing if Kaggle could host a competition by some African company or, even better, a data for good competition about any of the critical issues in Africa.
Which tools did you use?
I used Python and the common libraries for data analysis: Pandas, Matplotlib and seaborn. Using that, along with the storytelling skills I had gained thanks to my previous published kernels, I was able to come up with that notebook.
What have you taken away from this competition?
For me, the first thing that I took away is the satisfaction from being able to share with the Kaggle community, where Africans are clearly underrepresented, an article to get them interested in the state of AI there -- all while using the results of the survey rather than writing a regular blog post. The second thing is, of course, the many new tricks I picked up from other published notebooks during the competition. I mean, the visualizations in many kernels were just off the chart. If I had to pick up a couple that I would recommend, I’d probably go with Heads or Tails’ (as always) and Andre Sionek’s.
Mhamed Jabri is a MS student in Applied Mathematics at the Ecole Centrale Lyon in France. He aims to pursue a PhD, and his research interests include humanitarian AI and applications of ML to healthcare.