Train Delay Estimation

Overview

• All the statistical analysis (EDA, Data Preparation, model) are in the Solution.ipynb notebook.
• Additionally you can also refer to "report.html" for the EDA.
• Please find the RishabPal_solution.csv file for the test results.

Environment

• python == 3.6.8
• Ubuntu 18.04 LTS

Setup

$ pip install -r requirements.txt

Evaluation Results (on 10% validation set)

Random Forest Regressor

• ArrivalDelay MAE: 0.451

• DepartureDelay MAE: 0.043

• ArrivalDelay RMSE: 23.662

• DepartureDelay RMSE: 23.107

LightGBM

• ArrivalDelay MAE: 0.109

• DepartureDelay MAE: 0.489

• ArrivalDelay RMSE: 36.951

• DepartureDelay RMSE: 36.348

Improvements

• Identify more Key Performace Indocators (KPIs) and add it to the dataset.
• Data Distribution between train, validation and test set should be kept uniform.
• Train on a larger dataset.
• Use time-series analysis.

Papers

• Estimating Train Delays in a Large Rail Network Using a Zero Shot Markov Model
• Predictive Model of Train Delays in a Railway System