Improve your skills while having lots of fun
Data Science Hackatons are a great way to learn and improve your skills while having lots of fun. During a hackaton you join a team to compete with others to solve a data related puzzle in a limited time span. Recently, three TMC’ers independently joined the Data Science Hackaton organized by Itility in the Eindhoven area. Two of them ended up joining the winning team!
The event started off with a good dinner and some time to network with the participants. Since this hackaton lasts for only one evening, we then quickly continued with an introduction of the problem. One of the best parts of this hackaton was that we had to work with data from an actual customer of Itility, and the client was present during the event to explain the goal.
In short, it was an IT related problem, where the client had to license various software applications running on several servers across the world. The cost changed depending on where the application was run. For example, running an application on the global server was more expensive than on a local server, however, on the global server multiple users could share a single license. The task was to minimize the cost of all the licenses, while maximizing the availability for the users. For this we would be awarded points. Additional bonus points were awarded for any anomalies or curiosities found in the data.
After the introduction, we formed ad-hoc teams of four people, practically with whomever you sat next to at that moment. I thus teamed up with two Itility employees and Joris, who is also from TMC. The third person from TMC, Josine, ended up in another team. We acquired the data, did some quick brainstorming and visualizations, and decided we needed a certain metric to start the optimization. This proved harder than imagined and we spent almost all our time calculating this metric, until cracking it 10 minutes before the end. Then there was a stressful period in which we forced the solution in the right format and made a presentation out of our ideas and findings. We thought we had no way we could beat 6 other teams with this much chaos!
However, when the final points were calculated, we had an excellent score. We really amazed ourselves and the client with this achievement. In conclusion, if you like data science or are curious about it, I can really advise such a Hackaton. So who will join us for the next one?
Written by: Matthijs Cox