plkmo / GrabChallenge-Safety

Classification of dangerous driving based on telemetrics data

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

GrabChallenge-Safety

Instructions for inferring test set:

  1. Clone this repo
  2. copy test_set file (.csv format) into submission_data folder
  3. open command prompt at src folder
  4. run command: python model_infer2.py test_file.csv

Requirements: torch==1.0.0 sklearn==0.0 pandas==0.23.4 numpy==1.15.4 matplotlib==2.2.3

Training

Perform 5-fold split for validation testing alt text Loss vs Epoch alt text Training AUC vs Epoch alt text Test AUC vs Epoch alt text Test Accuracy vs Epoch

Model starts to overfit after 40 epochs. For final model, all datasets are included for training and early stopping is implemented after 30 epochs.

About

Classification of dangerous driving based on telemetrics data


Languages

Language:Python 100.0%