This week we held our workshop at NN, thank you for the venue!
First, Miklós Doma held a demo about solving a problem via training a Random Forest Classifier. We talked about the importance of feature selection, mean absolute error as a performance measure and how to visualize the decision tree itself.
After a short break, Hajnalka Kristóffy, continuing the proud tradition of working with the Pump it Up dataset, explained us the mathematical foundations of the Support Vector Machine approach with some stunning visualisations. We heard about the difference between linear and quadratic problems, saddle points and how to evaluate the models’ decision function.
We are very happy that we filled the provided room this time, had exciting discussions during/between and after the presentations and saw lively discussions all around the room.
Our next workshop will be hosted by EPAM, follow our social media channels for the exact details.
Thanks for everyone for attending!
As always, the notebooks are on our GitHub:
You can join us on our meetup page: