aditj / covertoptimization Goto Github PK

The repository is organized as follows:

• main_experiment/: Directory to store files for the main experiment in the paper. Contains a README file with instructions to run the code and preprocess the dataset
• bandit_benchmark_experiment/: Directory containing files to run the bandit benchmark experiment. Contains a README explaining the different files
• archive/: Directory to store old and misc (non-related) files; Does not have a README file.
• supplementary.pdf: Supplementary Document for the paper

The code is written in Python 3.7. The required libraries are:

• numpy
• pandas
• torch
• sklearn
• tqdm
• matplotlib
• pymdptoolbox (for the bandit benchmark experiment, specifically the backward induction algorithm) Initialize the following folders:
• data/: Directory to store the raw, cleaned, and client datasets
• • client_datasets/
• parameters/: Directory to store the parameters for the different experiments
• plots/

Link to Hatespeech Dataset: https://www.kaggle.com/c/jigsaw-toxic-comment-classification-challenge