Publications.

Papers, preprints, and scholarly work.

Publication index

2026

TAFG-MAN: Timestep-Adaptive Frequency-Gated Latent Diffusion for Efficient and High-Quality Low-Dose CT Image Denoising

Tangtangfang Fang, Yang Jiao, Xiangjian He, Jingxi Hu, Jiaqi Yang

Preprint · arXiv:2603.20868 · 2026

Timestep-adaptive frequency guidance for latent diffusion denoising, balancing noise suppression and anatomical detail preservation in low-dose CT.

Transforming Credit Risk Analysis: A Time-Series-Driven ResE-BiLSTM Framework for Post-Loan Default Detection

Yue Yang, Yuxiang Lin, Ying Zhang, Zihan Su, Chang Chuan Goh, Tangtangfang Fang, Anthony Graham Bellotti, Boon Giin Lee

Journal article · Information, 17(1), 5 · 2026

Combines attention and residual temporal modeling to capture repayment dynamics for post-loan default prediction.

2025

MAN: Latent Diffusion Enhanced Multistage Anti-Noise Network for Efficient and High-Quality Low-Dose CT Image Denoising

Tangtangfang Fang, Jingxi Hu, Xiangjian He, Jiaqi Yang

Preprint · arXiv:2509.23603 · 2025

Low-dose CT denoising in a compact latent space, studying the trade-off between image quality and diffusion-model inference cost.

A comprehensive study on the interplay between dataset characteristics and oversampling methods

Yue Yang, Tangtangfang Fang, Jinyang Hu, Chang Chuan Goh, Honghao Zhang, Yongmei Cai, Anthony Graham Bellotti, Boon Giin Lee, Zhong Ming

Journal article · Journal of the Operational Research Society, 76(10), 1981–2002 · 2025

Compares oversampling–classifier combinations across 16 datasets to examine how data characteristics shape predictive performance.

Under review

Learning Calibrated and Disentangled Anomaly Profiles for Time-Series Anomaly Detection under Missing Type Annotations

Yue Yang, Tangtangfang Fang, Howard Gerrard Zhuang, Chen Zhang

Under review at AAAI 2027 (second round).

Uses normal-reference calibration and intervention-guided disentanglement to characterize anomaly patterns when type annotations are scarce.