AutoEncoder Inspired Unsupervised Feature Selection

TL;DR

This paper introduces AEFS, an unsupervised feature selection method using autoencoders and group lasso to identify important features by capturing both linear and nonlinear relationships, improving performance on high-dimensional data.

Contribution

The paper presents a novel autoencoder-based feature selector that combines regression and group lasso, enabling more flexible unsupervised feature selection beyond linear assumptions.

Findings

AEFS outperforms state-of-the-art methods on benchmark datasets.

It effectively captures both linear and nonlinear feature relationships.

Experimental results demonstrate superior feature selection performance.

Abstract

High-dimensional data in many areas such as computer vision and machine learning tasks brings in computational and analytical difficulty. Feature selection which selects a subset from observed features is a widely used approach for improving performance and effectiveness of machine learning models with high-dimensional data. In this paper, we propose a novel AutoEncoder Feature Selector (AEFS) for unsupervised feature selection which combines autoencoder regression and group lasso tasks. Compared to traditional feature selection methods, AEFS can select the most important features by excavating both linear and nonlinear information among features, which is more flexible than the conventional self-representation method for unsupervised feature selection with only linear assumptions. Experimental results on benchmark dataset show that the proposed method is superior to the state-of-the-art method.