by: Xavier M. Puspus
Email: xpuspus@gmail.com
Country: Philippines

Given the telematics data for each trip and the label if the trip is tagged as dangerous driving, derive a model that can detect dangerous driving trips.

The given dataset contains telematics data during trips (bookingID). Each trip will be assigned with label 1 or 0 in a separate label file to indicate dangerous driving. Pls take note that dangerous drivings are labelled per trip, while each trip could contain thousands of telematics data points. participants are supposed to create the features based on the telematics data before training models.

The notebook safety_challenge_xmpuspus.ipynb is the main notebook for the challenge. It contains codes for loading the data, feature engineering, model training, model scoring and hold-out scoring. To run the notebook, please place the dataset in the data/safety/ folder in their corresponding directories.

The folder data/ contains all data used for this challenge. It contains the safety folder as downloaded from the grab AIforSEA Safety Challenge. The folder model/ contains the pickled trained model should the examiner choose to just measure on holdout set. Please see last section of this ReadMe for details on how to model on holdout set.

I trained a classifier on the first 120 seconds of each feature/signal, concatenated into a single array and fed it into a feed-forward multi-layer neural network. Details of the data pre-process is in the notebook stated above as well as in the utils/utils.py script in the function process_data().

To examiner, please save hold out data to data/test/ folder with the same folder structure as the one provided for the challenge in safety/ folder (hold out features should be in a folder named features/ and the labels should be in a folder labels/, where both folders should be placed in data/test/ folder). Run cells below the header Measure From Holdout Data once the holdout data is in the suggested folder.