A package of distributionally robust optimization (DRO) methods. Implemented via cvxpy and PyTorch

[NeurIPS2023] Official code of "Understanding Contrastive Learning via Distributionally Robust Optimization"

The Pytorch implementation for "Topology-aware Robust Optimization for Out-of-Distribution Generalization" (ICLR 2023)

"Aligning Distributionally Robust Optimization with Practical Deep Learning Needs"

Distributionally robust machine learning with Pytorch and Scikit-learn wrappers

[ICLR 2025] Official code of "Towards Robust Alignment of Language Models: Distributionally Robustifying Direct Preference Optimization"

Temporally and Distributionally Robust Optimization for Cold-start Recommendation (AAAI'24)

[ICDE2024] Official code of "BSL: Understanding and Improving Softmax Loss for Recommendation"

This is the official repository for the ICLR 2024 paper Out-Of-Domain Unlabeled Data Improves Generalization.

Python Implementation of the Instance-wise Distributionally Robust Nonnegative Matrix Factorization (iDRNMF)

Python Implementation of DRNMF-SP

Code for the experiments in the paper "Contextual Robust Optimisation with Uncertainty Quantification".

Code for the project DiGriFlex

An open-source Python module for portfolio optimization and backtesting

R and Python demos for paper `Robust Mixture Models for Algorithmic Fairness Under Latent Heterogeneity'

Code for the Paper "Robust Offline Reinforcement Learning with Linearly Structured f-Divergence Regularization", International Conference on Machine Learning (ICML) 2025

this repository hosts my master's thesis codes and dataset.

End-to-end Python framework for robust pension fund management under parameter uncertainty. Implements three Distributionally Robust (DRO) Asset-Liability Management (ALM) formulations (Mixture, Box, Wasserstein) with GBM scenario generation, convex optimization (LP/SOCP), and comprehensive backtesting. Based on 2026 research by Ghahtarani et al.

Distributionally robust Cox regression with Wasserstein ambiguity sets. Includes classical baselines, evaluation metrics, and reproducible experiments.

End-to-End Python implementation of Liu & Cheng's (2026) methodology for U.S. Treasury yield curve forecasting. Combines Factor-Augmented Dynamic Nelson-Siegel models, High-Dimensional Random Forests, and Distributionally Robust Optimization (DRO) for risk-aware ensemble forecasting under ambiguity.