Having hands-on experience is critical when wanting to launch a new career as a data scientist. We are therefore constantly on the lookout for current and practical student projects to put the knowledge of our juniors to the test. This time, we teamed up with IBM to offer our Liège and Charleroi students the unique opportunity of diving into Covid data. Curious about what IBM had in store for them or how they’ve experienced this collaboration? Then quickly read the article below!
The most difficult part of kicking off a career as a data scientist is undoubtedly the beginning. Most open vacancies require a minimum of 3 years of experience with technologies such as machine learning, deep learning or computer vision. Even though online training paths and courses in traditional education may help you gain the necessary knowledge, knowing how to put this knowledge into practice is a totally different story. After having followed such training courses, the question inevitably arises: how can you be qualified for a job in the field without having any concrete experience?
We strongly believed that it could be done differently. And so our active pedagogy was born. At BeCode, our students learn by doing, by applying their skills to concrete projects. The learning environment can be compared to a company’s office, which gives them a taste of what a job as a data scientist entails. However, closing this skills gap is a challenge that we can only tackle thanks to our strong network of partners. To provide our Charleroi and Liège students with a new, current challenge, we teamed up with IBM for a three-part workshop session on exploring, visualising and modelling Covid data.
Tackling the healthcare crisis by means of AI
Damiaan Zwietering, developer advocate at IBM, functions as an intermediary between IBM and the developer community. By organizing tech talks and workshops he aims at inspiring people to kick off a career in AI or at helping those already active in the field to discover new techniques and technologies.
And as experience has taught him, there is no better source of inspiration for these kinds of sessions than your own daily life. The prevailing health crisis is significantly increasing the pressure on the healthcare industry, a phenomenon Damiaan has witnessed with his own eyes as his wife is a nurse. In the hope of contributing to the suppression of this crisis, Damiaan analysed and interpreted the current figures and started organizing workshops around his findings, so too for our students.
Getting ready for a career as a data scientist
The workshop, consisting of three sessions, aimed at immersing our students in the wonderful world of data science. “This project represents the journey that the BeCode juniors will take when working on a project on behalf of a client, and that’s exactly why I wanted to give this workshop to them. I aimed at introducing them to the different stages of a client-oriented project: researching, performing analyses, interpreting results, building models, visualizing the outcome etc.”, explains Damiaan. And this, of course, with the necessary guidance and support. “During the first session, I presented different datasets, tools and interesting avenues of research. Afterwards, the students brainstormed on the research question, related to covid, that they would like to answer. These were reviewed and discussed during a second session, a Q&A”, Damiaan continues.
However, coming up with the right research question wasn’t a piece of cake. “As Damiaan had worked on this topic for over a year now, he had already answered lots of different research questions. It was therefore difficult to come up with a new approach, but we succeeded. We decided to test whether Trump supporters had a bigger chance of getting Covid”, explains BeCode Liège junior Guillaume.
Morgane’s group was particularly ambitious and decided to even tackle two research questions, focusing on the severity of covid and the number of bankruptcies as a result of this crisis. And the results were surprising. “We had expected that many more failures could be attributed to the corona crisis, but the government measures have prevented more companies from going bankrupt than normally. Yet, it’s important to not draw any final conclusions as the crisis isn’t over yet”, explains BeCode Charleroi junior Morgane.
But not only finding the right research question, also working remotely in a team with people that you’ve never met before can be a challenge. “Each group consisted of three people, two learners from Liège, one from Charleroi. It was really interesting to work with people that you usually don’t work with as it offers you the opportunity of gathering another point of view”, explains Morgane. “But working on such projects remotely, is a challenge. Luckily it wasn’t the first time that we had to collaborate online, so we’re getting better at it”, she continues.
BeCode wouldn’t be BeCode if our juniors didn’t have to present the final result. “The presentations went pretty well. Damiaan was happy with and surprised by the results of our analysis. We discussed our approach and he provided us with constructive feedback and leads on interesting follow up analyses”, says Guillaume.
And let reviewing the results with a critical eye be exactly what Damiaan considers important. “I’m used to questioning and discussing the obtained results. Curiosity is one of the most important qualities of a data scientist. You should never open a dataset and immediately perform an analysis on it without asking yourself first where the data comes from. Next to that, you should also be critical towards your obtained results. Even if you get great results, you should always ask yourself why. That’s where the word humble comes in. You must never wave your results around as if they were the only truth, they are rather an expression of how you see things”, thus Damiaan.
That’s also the lesson that Morgane has learned via these workshops. “It’s important to do research first and to be careful when interpreting the results of your analysis. Damiaan has taught us not to draw up conclusions immediately and to bear in mind that we’ve based our analysis on a dataset that we do not always know the origin of.”
Also for Damiaan, these workshops were very instructive and certainly worth repeating. “This workshop was also very inspiring for me as I’ve learned lots of new things. No major insights, mainly a confirmation of how things work and of what challenges people are facing”, thus Damiaan. “I would love to do this again, so please feel free to ask. I really enjoyed giving these workshops and as for you guys, keep up the good work. You’ve a bunch of amazing teachers and I hope you can soft-land your juniors in their new career path, but I know that you’re doing everything in your power to do so”, he concludes.
Interested in following our AI bootcamp?
Fantastic! – There are new classes starting in Ghent, Antwerp, Brussels and Liège.