Football_Match_Probability_Prediction

The project is built for educational purposes, to learn how to apply feature engineering techniques, data preprocessing, model selection, and paramters tuning applied on big data. This project is based on a Kaggle competition you can find its link here: https://www.kaggle.com/competitions/football-match-probability-prediction

1. Predicted the probabilities of more than 150000 match outcomes using the recent sequence of 10 matches of the teams with over 180 features.
2. Performed Market Basket Analysis to identify associations between features and look for combinations of items that occur together.
3. Visualized processed data to grasp some insights from them.
4. Tuned model to select the best output on our metric RMSLE-Root Mean Squared Log Error.