Evo-TFS: Evolutionary Time-Frequency Domain-Based Synthetic Minority Oversampling Approach to Imbalanced Time Series Classification

2026-01-03

arXiv.org

AI Summary

This study addresses the challenge of class imbalance in time series classification, where models often overlook minority classes. To mitigate this issue, the authors propose an evolutionary oversampling approach that integrates time-domain and frequency-domain features. For the first time, time–frequency joint features are incorporated into a strongly typed genetic programming framework. A novel fitness function is designed to simultaneously capture temporal dynamics and spectral characteristics, enabling the generation of high-quality, diverse synthetic samples for minority classes. This strategy overcomes the limitations of conventional linear interpolation methods, which struggle to preserve both structural integrity and diversity in time series data. Experimental results across multiple imbalanced time series datasets demonstrate that the proposed method significantly outperforms existing oversampling techniques and effectively enhances the performance of both time-domain and frequency-domain classifiers.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Search and Optimization: Evolutionary Computation
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Web Mining and Content Analysis: Models for Web evolution
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Time series classification is a fundamental machine learning task with broad real-world applications. Although many deep learning methods have proven effective in learning time-series data for classification, they were originally developed under the assumption of balanced data distributions. Once data distribution is uneven, these methods tend to ignore the minority class that is typically of higher practical significance. Oversampling methods have been designed to address this by generating minority-class samples, but their reliance on linear interpolation often hampers the preservation of temporal dynamics and the generation of diverse samples. Therefore, in this paper, we propose Evo-TFS, a novel evolutionary oversampling method that integrates both time- and frequency-domain characteristics. In Evo-TFS, strongly typed genetic programming is employed to evolve diverse, high-quality time series, guided by a fitness function that incorporates both time-domain and frequency-domain characteristics. Experiments conducted on imbalanced time series datasets demonstrate that Evo-TFS outperforms existing oversampling methods, significantly enhancing the performance of time-domain and frequency-domain classifiers.

Problem

• imbalanced time series classification
• minority oversampling
• temporal dynamics
• sample diversity
• time-frequency domain

Innovation

• evolutionary oversampling
• time-frequency domain
• genetic programming
• imbalanced time series classification
• synthetic minority oversampling