Potential Anchoring for imbalanced data classification

TL;DR

This paper introduces Potential Anchoring, a novel resampling framework using radial basis functions to improve classification on imbalanced datasets by preserving class distribution shapes and optimizing synthetic data placement.

Contribution

It presents a unified over- and undersampling method that outperforms existing algorithms, especially on naturally complex, noise-free datasets.

Findings

Outperforms state-of-the-art resampling algorithms on 60 datasets

Effectively preserves class distribution shapes during resampling

Excels on naturally complex, noise-free datasets

Abstract

Data imbalance remains one of the factors negatively affecting the performance of contemporary machine learning algorithms. One of the most common approaches to reducing the negative impact of data imbalance is preprocessing the original dataset with data-level strategies. In this paper we propose a unified framework for imbalanced data over- and undersampling. The proposed approach utilizes radial basis functions to preserve the original shape of the underlying class distributions during the resampling process. This is done by optimizing the positions of generated synthetic observations with respect to the potential resemblance loss. The final Potential Anchoring algorithm combines over- and undersampling within the proposed framework. The results of the experiments conducted on 60 imbalanced datasets show outperformance of Potential Anchoring over state-of-the-art resampling algorithms, including previously proposed methods that utilize radial basis functions to model class potential. Furthermore, the results of the analysis based on the proposed data complexity index show that Potential Anchoring is particularly well suited for handling naturally complex (i.e. not affected by the presence of noise) datasets.