GAIS: A Novel Approach to Instance Selection with Graph Attention Networks

TL;DR

GAIS is a new graph attention network-based method for instance selection that effectively reduces dataset size while preserving or enhancing model accuracy across diverse datasets.

Contribution

Introduces GAIS, a novel graph attention network approach for instance selection that captures complex data relationships and outperforms traditional methods.

Findings

Achieves an average data reduction of 96% while maintaining accuracy.

Outperforms traditional instance selection methods across 13 datasets.

Slightly higher computational cost but better data efficiency.

Abstract

Instance selection (IS) is a crucial technique in machine learning that aims to reduce dataset size while maintaining model performance. This paper introduces a novel method called Graph Attention-based Instance Selection (GAIS), which leverages Graph Attention Networks (GATs) to identify the most informative instances in a dataset. GAIS represents the data as a graph and uses GATs to learn node representations, enabling it to capture complex relationships between instances. The method processes data in chunks, applies random masking and similarity thresholding during graph construction, and selects instances based on confidence scores from the trained GAT model. Experiments on 13 diverse datasets demonstrate that GAIS consistently outperforms traditional IS methods in terms of effectiveness, achieving high reduction rates (average 96\%) while maintaining or improving model performance. Although GAIS exhibits slightly higher computational costs, its superior performance in maintaining accuracy with significantly reduced training data makes it a promising approach for graph-based data selection.