musamawaheed / Project4_BoschProductionPerformancePrediction

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Eight code files are provided for this project. The results shown in the report can be reproduced by using those files in combination with the data (available at https://www.kaggle.com/c/bosch-production-line-performance/data) and the parameter-settings explained in the text.

Three files show the applied feature selection methods.
- feature_selection_modelbased.py
- feature_selection_univariate.py
- feature_selection_xgb.py

In same cases the feature selection files were taken to create new datasets that were then used by the models.

The files for the models are:
- isolation_forest.py
- random_forest_numeric.py
- random_forest_numeric_date.py
- xgb_numeric.py
- xgb_numeric_date.py

About


Languages

Language:Python 100.0%