Skip to content

Latest commit

 

History

History
executable file
·
19 lines (13 loc) · 949 Bytes

README.md

File metadata and controls

executable file
·
19 lines (13 loc) · 949 Bytes

MAP3.0

This project was coded in a modular format with utility scripts for ease of mainenance, understandability and reusability. For the ability to execute them in one shot, all the utility scripts were copied to final.ipynb.

Steps to reproduce submission file on Kaggle

  • Assuming the dataset is present in Kaggle environment/kaggle/input *
  • Upload notebook in final.ipynb to Kaggle
  • Execute all cells sequentially
  • Prediction file with timestamp can be found in /kaggle/output directory

Steps to reproduce submission file locally

  • Update PROJECT_HOME variable to point root of project
  • Download and store the train, test data in the directory $PROJECT_HOME/input/fall2021-inf8245e-machine-learning as x_train.pkl, y_train.pkl, x_test.pkl
  • Notebook is present in $PROJECT_HOME/code/final.ipynb
  • Execute all cells sequentially
  • Prediction file with timestamp can be found in $PROJECT_HOME/output directory