Regression Augmentation With Data-Driven Segmentation

2025-08-02

AI Summary

In regression tasks, skewed target distributions cause imbalance—models overfit dense regions while underrepresenting sparse regions with few samples. To address this, we propose a fully data-driven augmentation framework. Methodologically, we introduce a novel Mahalanobis-Gaussian Mixture Model (Mahalanobis-GMM) that adaptively identifies genuine sparse regions in the joint feature–target space without relying on manually defined thresholds. We further combine deterministic nearest-neighbor matching with generative adversarial network (GAN)-based synthesis to precisely augment rare samples. Extensive experiments across 32 benchmark regression datasets demonstrate that our approach significantly outperforms state-of-the-art data augmentation methods. Notably, it substantially improves generalization and prediction robustness for minority samples—those residing in low-density target regions—while maintaining accuracy across the full distribution.

Technology Category

• Machine Learning: Feature Construction/Reformulation
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Imbalanced regression arises when the target distribution is skewed, causing models to focus on dense regions and struggle with underrepresented (minority) samples. Despite its relevance across many applications, few methods have been designed specifically for this challenge. Existing approaches often rely on fixed, ad hoc thresholds to label samples as rare or common, overlooking the continuous complexity of the joint feature-target space and fail to represent the true underlying rare regions. To address these limitations, we propose a fully data-driven GAN-based augmentation framework that uses Mahalanobis-Gaussian Mixture Modeling (GMM) to automatically identify minority samples and employs deterministic nearest-neighbour matching to enrich sparse regions. Rather than preset thresholds, our method lets the data determine which observations are truly rare. Evaluation on 32 benchmark imbalanced regression datasets demonstrates that our approach consistently outperforms state-of-the-art data augmentation methods.

Problem

• Addresses imbalanced regression with skewed target distributions
• Identifies minority samples without fixed ad hoc thresholds
• Enhances sparse regions using data-driven GAN-based augmentation

Innovation

• GAN-based augmentation for imbalanced regression
• Mahalanobis-GMM identifies minority samples automatically
• Deterministic nearest-neighbour matching enriches sparse regions