WiDS Datathon event in Qatar was held for two days, the 30th of January and the 6th of February, organized by Hamad Bin Khalifa University (HBKU) and the WiDS Qatar 202 Ambassadors Halima A, Rachael Fernandez, Reem Suwaileh & Maram Hasanain.
The goal of the first event was to introduce Kaggle and refresh basic Machine Learning (ML) concepts. The second event was a coding marathon, where we provided an informative, fun and collaborative environment for the participants to build a team of datathoners!
First workshop: Machine Learning 101
The first workshop was held on Thursday 30th January 2020 from 4:00 pm – 8:00 pm in Lecture Hall A, LAS Building, Education City, Qatar.
We created a promotion on Instagram in an effort to reach out to more participants. We ran a total of 2 promotions, 1 for each workshop and were able to reach about 10k people living in Doha primarily between the ages of 18-34.
The event agenda was prepared to make sure the attendees were introduced to all the concepts that are required to take part in the Datathon. Around 70 participants attended the workshop out of 100 registrants from diverse backgrounds including female and male students (graduate students, postgrads), computer engineers, fresh graduates, statisticians, postdocs, research assistants, senior scientists, QoS scientists, etc.
The first workshop began with an introduction to WiDS and some highlights about previous WiDS Qatar conferences. The first session - “Intro to ML” introduced the participants to the fundamental concepts of ML. Some of the covered topics include: ML and its applications, Supervised vs Unsupervised learning, ML Models and Performance measures. Detailed agenda is below:
4: 00 - 4: 20 [20 min] Registration and refreshments
4: 20 - 5: 20 [60 min] Intro ML
5: 20 - 5: 40 [20 min] Break (prayer 5:18pm)
5: 40 - 6: 10 [30 min] Intro to Kaggle
6: 10 - 6: 20 [10 min] Break
6: 20 - 7: 10 [50 min] Hands-on
7: 10 - 7: 30 [20 min] Summary and overview of the next session
Participants were then introduced to the Kaggle platform and were taken through the process of creating a notebook, making submissions and viewing their standing on the leaderboard. This was followed by a hands-on session in which the participants were given hands-on training on one of the Python notebooks from the WiDS tutorials.
We finally ended with key takeaways and a sneak-peek into what can be expected for the next workshop.
Second workshop: Coding Marathon!
The second workshop was held on Thursday 6th February 2020 from 4:00 pm – 8:00 pm in Lecture Hall A, LAS Building, Education City, Qatar.
We prepared an informative, fun and collaborative environment to build a team of datathoners and hone participants’ data science skills!
Around 15 (11 female and 4 male) participants attended the workshop with different backgrounds.
The workshop started with a quick introduction to the WiDS and the Datathon followed by an ice-breaker session that helped attendees to get to know each other before forming teams.
4: 00 - 4: 20 [20 min] Registration and refreshments
4: 20 - 4: 35 [15 min] Overview of the marathon
4: 35 - 5:00 [25 min] Forming teams
5: 00 - 5: 30 [30 min] Preparing competition environment
5: 30 - 5: 50 [20 min] Break (prayer 5:18pm)
5: 50 - 6: 20 [30 min] Team brainstorming
6: 20 - 7: 35 [75 min] Competition
7: 35 - 7:50 [15 min] Closing
We started workshop 2 by giving the participants an overview of the Datathon, the rules for participating and the type of dataset that they will be tackling. The participants were then encouraged to form teams and we had a short ice-breaker to help them to scout for team members. At the end of this workshop, we had around 5 teams ready to compete! The teams were then encouraged to find common ground in the language they would use (R or Python) and other collaborative tools. After the break, the teams brainstormed ideas to tackle the Datathon within their newly formed teams, after which they started to work on the Datathon.
Overall, the two workshops were fun! We had a full-house in the first workshop and we had a diverse set of participants in terms of their ML background. Their expertise ranged from working in ML jobs to students who were taking ML courses to people who were new to the whole idea of ML! The second workshop had fewer participants which we anticipated given that most of the participants from workshop 1 were beginners. However, we had a very enthusiastic set of participants in workshop 2 who were highly motivated to take part in the challenge.
This was the first time that we held workshops before the main conference and identified some shortcomings. We would like to find a way to tackle the different ML backgrounds of the participants, check if weekend workshops would attract more participation and possibly invite ML professors / ML practitioners to give some of the workshops in the future.
Follow us on Twitter @WiDS_Qatar
Follow us on Instagram @wids.qatar
Send us an email: wids_qa@gmail.com