Arin Chang

I am a PhD student in Statistics at Purdue University, advised by Jordan Awan and Vinayak Rao. My research lies at the intersection of differential privacy, computational statistics, and machine learning, with a particular focus on developing statistically valid and computationally efficient methods for inference from privatized data. My current work includes Bayesian computation and MCMC for private data, simulation-based inference, and privacy-preserving statistical methodology. More broadly, I am interested in privacy-preserving machine learning, statistical inference, and computational methods for modern data analysis.

Email: chan1074 [at] purdue [dot] edu

selected publications

TMLR

Hamiltonian Monte Carlo for Bayesian Inference from Privatized Data

Arin Chang, Jordan Awan, and Vinayak Rao

In submission to Transactions on Machine Learning Research (TMLR). Preprint coming soon, 2026

JMLR

Optimal Debiased Inference on Privatized Data via Indirect Estimation and Parametric Bootstrap

Zhanyu Wang, Arin Chang, and Jordan Awan

Under review at Journal of Machine Learning Research (JMLR), 2026

arXiv Bib HTML

@article{wang2025debiased,
  title = {Optimal Debiased Inference on Privatized Data via Indirect Estimation and Parametric Bootstrap},
  author = {Wang, Zhanyu and Chang, Arin and Awan, Jordan},
  journal = {Under review at Journal of Machine Learning Research (JMLR)},
  year = {2026},
}

ICML

A Three-regime Model of Network Pruning

Yefan Zhou, Yaoqing Yang, Arin Chang, and 1 more author

In International Conference on Machine Learning (ICML), 2023

arXiv Bib HTML

@inproceedings{zhou2023three,
  title = {A Three-regime Model of Network Pruning},
  author = {Zhou, Yefan and Yang, Yaoqing and Chang, Arin and Mahoney, Michael W.},
  booktitle = {International Conference on Machine Learning (ICML)},
  year = {2023},
}