Analysing data and deciding which unnecessary features to drop
Preprocessing the train data and evaluate different models (using 70% of train data for training and the rest 30% for testing the model) and compared their r2 score and mae to choose the best. as a result, the ensemble model RandomForestRegressor is the best model for us.
Model | R2 Score | MAE | CV R2-Score |
---|---|---|---|
RandomForestRegressor | 0.9006 | 0.9263 | 0.8946 (+/-) 0.0015 |
Preprocessing train and test data, training the RandomForestRegressor model on train data and then predict the test data.
the final prediction is saved in test_prediction.csv file.
- sklearn
- pandas
- numpy
- matplotlib