GECCO-2020-PyTorch

A Genetic Algorithm to Optimize SMOTE and GAN Ratios in Class Imbalanced Datasets.

The experimental codes using PyTorch from the paper that was submitted to GECCO 2020. (https://doi.org/10.1145/3377929.3398153)

Getting Started

Environments

• Ubuntu 16.04 or 18.04
• Python 3.6 or 3.7

Installation from PyPi

• PyTorch 1.4.0
• scikit-learn
• pandas
• DEAP
• imbalanced-learn

Installation from Docker

• Dockerfile

Codes

• classifier/ : A python module implementing NN-based classifier.

• ga/ : A python module that implements the GA method to find the optimal oversampling ratio.

• gan/ : A python module implementing GAN-based sampling method.

• oversample.py : Executable script to oversample minority data using SMOTE, SVMSMOTE, GAN, etc.

• train.py

• eval.py

• search_GA.py

• train_GA.py