aiqbench analysis
This research introduces Quantile-TabDDPM, a diffusion model enhanced with quantile regularization to better handle extreme class imbalances in tabular data like credit card fraud detection. By modifying the denoising objective to capture rare events alongside standard loss functions, the method addresses a key limitation of current generative approaches when dealing with heavy-tailed distributions. The study validates this framework on real-world datasets, demonstrating its potential to improve minority-class performance without sacrificing theoretical grounding.
Usefulness of Quantile-Aware Diffusion Modeling for Highly Imbalanced Tabular Data
arXiv:2610.05825v1 Announce Type: new Abstract: Classification problem in the context of highly imbalanced data is a major challenge in many real-world applications (e.g., FinTech, healthcare, etc.). In these cases, the vast majority of instances belong to a single class and a small fraction represent the minority class (often the most critical class). Recently, diffusion models have emerged as powerful approaches to reduce the degree of ``imbalanced-ness'' in the dataset; they work by generating synthetic data by capturing complex data distributions using iterative transformations. However, standard diffusion models are not inherently suited to highly skewed or heavy-tailed data, due to inbuilt quadratic error loss, which lacks the structural sensitivity to capture rare, extreme values, and minority-class nuances. We propose a novel approach, namely, Quantile-TabDDPM, based on a quantile-regularized denoising objective that combines the standard quadratic error loss with a quantile loss term to explicitly capture rare events while preserving the theoretical grounding of the original denoising objective. We extensively evaluated our approach on a real-world credit card transaction dataset characterized by extreme class imbalance. The results demonstrate that the integration of diffusion-based synthetic data generation with a quantile-regularized denoising objective provides a robust and effective framework for fraud detection in highly imbalanced datasets.
Abstract:Classification problem in the context of highly imbalanced data is a major challenge in many real-world applications (e.g., FinTech, healthcare, etc.). In these cases, the vast majority of instances belong to a single class and a small fraction represent the minority class (often the most critical class). Recently, diffusion models have emerged as powerful approaches to reduce the degree of ``imbalanced-ness'' in the dataset; they work by generating synthetic data by capturing complex data distributions using iterative transformations. However, standard diffusion models are not inherently suited to highly skewed or heavy-tailed data, due to inbuilt quadratic error loss, which lacks the structural sensitivity to capture rare, extreme values, and minority-class nuances. We propose a novel approach, namely, Quantile-TabDDPM, based on a quantile-regularized denoising objective that combines the standard quadratic error loss with a quantile loss term to explicitly capture rare events while preserving the theoretical grounding of the original denoising objective. We extensively evaluated our approach on a real-world credit card transaction dataset characterized by extreme class imbalance. The results demonstrate that the integration of diffusion-based synthetic data generation with a quantile-regularized denoising objective provides a robust and effective framework for fraud detection in highly imbalanced datasets.
Machine Learning (cs.LG); Applications (stat.AP)
arXiv:2610.05825 [cs.LG]
(or arXiv:2610.05825v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.05825 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 293 words
Related coverage
Handling Imbalanced Classification: What Works Better Than SMOTE
Analytics Vidhya · Jul 12, 2026
TabFORGE: A Tabular Foundation Model for Generative Modelling
arXiv · Oct 15, 2026
SynPre-FL: Synthetic data-driven pretraining integrated Federated Learning training framework
arXiv · Jul 23, 2026
Logit-Coordinate Generative Models for Mixed Continuous-Categorical Tabular Data
arXiv · Jul 28, 2026