Classification Model for predicting football matches outcomes in Italian Serie
This project has been done in 4 days, during the Disruptive Data Summer School in Viterbo, Italy. The major objective was to deal with a large amount of data and manage to build from scratch our own model that makes the prediction of matches outcomes in Italian Serie A.
This project has been done in Python, using different librairies especially the scikit learn one for testing and training the models.
We arrived to an accuracy of 52% with the Gaussian NB method, the best bookers can predict the outcomes 58% of time...