APS Sensor Fault Detection

The APS Sensor Fault Detection is a problem of predicting Failure of Scania Trucks APS to decrease cost of reparing and also hinder dangrous events. It is modeled by various Machine learning algorithms such as Random Forest, XGBoost, Decision Tree, KNN, Logistic Regression and the best model finalized. THe hyperparameter tuning was done by Optuna library. The model deplyed by docker.

Step 1 - Download dataset

To downlaod dataset run this command in terminal
wget https://archive.ics.uci.edu/ml/machine-learning-databases/00421/aps_failure_training_set.csv

Step 2 - Install the requirements

pip install -r requirements.txt
Add connecting information of your Mongodb and AWS in .env file

Step 3 - Run main.py file

python main.py