/GDR

Algorithm for interpretable dimension reduction via symbolic regression on genetic programming

Primary LanguagePython

GDR

Genetic Dimension Reduction - algorithm for interpretable dimension reduction for tabular data in problem of binary classification. This repository contains prototype version of the algorithm and code used to make benchmarks.

Developed and implemented during research work in Novosibirsk State University by Nikita Radeev.

Code is available for execution online on CodeOcean: https://codeocean.com/capsule/3280561/tree

Test Data

Datasets are not uploaded here but you can find them on OpenML (https://www.openml.org/) by task id:

Dataset OpenML task id
Bank Marketing 1461
Blood Transfusion 1464
Breast Cancer Wisconsin 1510
Credit-g 31
Diabetes 37
Hyperplane 43122
Ionosphere 59
Madelon 1485
Sonar 40
Bioresponse 4134
Christine 41142
Guillermo 41159