It is necessary to predict equipment failure.
Initial data:
Various equipment operation data and recorded failures
Goal:
Conduct exploratory data analysis, find relationships (it is possible to explain the dependence - without training), train model
Uploaded to the repository:
- Equipments-Failures
- Basic model (without training): ROC_AUC - 0.884
- Several modals were trained, the best one was selected, hyperparameters were selected: ROC_AUC - 0.972 XGBClassifier
