Skip to content

dps123/AIMarkerFinder

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

AIMarkerFinder

AIMarkerFinder is a computational tool designed for the analysis of high-dimensional biomedical data (e.g., metabolomic or genomic datasets) to identify minimal sets of informative biomarkers. The method combines a denoising autoencoder with an attention mechanism (DAE) and a Kolmogorov–Arnold Network (KAN) to perform feature selection and build interpretable classification models. This approach effectively addresses the "curse of dimensionality" and feature redundancy, enhancing both model accuracy and interpretability.

Preprint: https://www.preprints.org/manuscript/202511.1705


Installation

Requirements

  • Python ≥ 3.10
  • Poetry for dependency management

Setup

Install dependencies using Poetry:

poetry install

Quick Start

View all available arguments:

poetry run aimf -h

Basic command syntax:

poetry run aimf <input_file> [options]

Supported input formats: .csv, .tsv, .xls, .xlsx
Class labels: Specify with -CC; if omitted, all non-numeric columns are used by default.

Example usage with an input file example/test_data.xlsx:

poetry run aimf example/test_data.tsv -balance -CC Diagnose

About

Feature selection algorithm

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages