Data Science Club — 1st Meetup

There is a new Data Science Club in town and by town I mean Bratislava. Yesterday was its kickoff called Intro into data science and we’ve met in FIIT’s Turing lecture room.

So how was it?

About 50 people from business and academic environment attended which, in my opinion, is an excellent attendance for any “first of” — so obviously people are interested in Data Science in Bratislava. Now I just hope that the news about the hottest club in town will spread like a wildfire.

Okay, enough of me making cheesy puns and back to how it went.

To start it off Jozo Kovac told us that the point of Data science club is to bridge the divide between academic and business and ensured everyone that it’s not Exponea’s self promo, so other companies are very welcome as speakers and we are going to tackle topics as problem first which means that theory takes a second fiddle to an actual solutions to perceived problems and explained that data science is intrinsically multidisciplinary, yet no one can be an expert on everything.

Then Robert Lacok took over and his talk was about data science process and challenges that data scientists face which, to a certain degree, we’ve went through ourselves.

After Robo’s talk, our task was to go through the process of creating a prediction model on dataset from Kaggle and the guys from Exponea’s AI team were on hand to help us with whatever hiccup we might have had.

Even though I have some Python experience, I’m strongest in the domain and business expertise spectrum of data science, but with the help of everyone, I was able to run my first prediction model without any tool such as Exponea or Rapid Miner — wohoo.

And to finish up our meetup — PIZZA arrived O_O and chatting with newly unearthed fellow data science enthusiasts while munching on a slice of pizza was the best.

So to summarize it — I’m looking forward to the next meetup in about two weeks and if you’d like to join, follow this Facebook page.

Key takeaways:

• Data science club isn’t self promo event
• Data science is intrinsically multidisciplinary and no one can really be an expert on every aspect of it
• Problem first approach
• Creating a decision tree isn’t that hard

Slides from the meetup are here: Jozo’s slides, Robo’s slides

Data from the workshop are here: GitHub