This repository contains the optional homework to get additional points for the final exam.
This exam was inserted in my study plan as an external one from Robotics and AI (Professor Luca Iocchi).
As an optional task it was possible to make predictions on blind tests and send them as csv files (one for each dataset in the hw). Based on the predictions a position on the leadearboard was assigned.
Each homework contains the official requirements, my pdf report and the python code used to solve the tasks. I started using colab after the first homework since it's so much better.
Multiclass classification with any prefered model.
The only requirement was to compare 2 different models for each dataset.
Both datasets were balanced and had 5000 instances per class (10 total classes) for a total of 50k elements.
The elements of the datasets were feature vectors of dimension d, specifically 100 for dataset 1 and 1000 for dataset 2.
2/2
-/84
Due to malformatted files the ranking was not assigned for me (by following the description it had to have two columns, but it apparently only needed to have one).
Coming soon
?/2