Chester Holtz

University of California San Diego
San Diego, CA 92093
Email: chholtz AT ucsd.edu

CV: (long ver.), (short ver.)
Google scholar
Github
Student evaluations



I'm a Postdoc in HDSI at UCSD working with Gal Mishne. I'm interested in graphs and machine learning and optimization. I did my PhD with Chung-Kuan Cheng. Previously I interned as a quantitative researcher with the systematic options alpha group at Citadel Securities. I also like horror movies, swimming, and snowboarding and a bit of running.

Here are some papers from my PhD and Postdoc. Some of the authorship ordering is alphabetical

Dhruv Kohli, Jesse He, Chester Holtz, Gal Mishne, Alexander Cloninger, Robust estimation of boundary using doubly stochastic scaling of Gaussian kernel, arXiv:2411.18942.

Chester Holtz, Pengwen Chen, Zhengchao Wan, Chung-Kuan Cheng, Gal Mishne, Continuous Partitioning for Graph-Based Semi-Supervised Learning, arXiv:.

Pengwen Chen, Chung-Kuan Cheng, Chester Holtz, Sequential subspace methods on Stiefel manifold optimization problems, arXiv:2404.13301.

Chester Holtz, Pengwen Chen, Alexander Cloninger, Chung-Kuan Cheng, Gal Mishne, Semi-Supervised Learning on Stiefel Manifolds, arXiv:2308.00142.

Chester Holtz, Yucheng Wang, Chung-Kuan Cheng, Bill Lin, On Robustness and Generalization of ML-Based Congestion Predictors to Valid and Imperceptible Perturbations, arXiv:2403.00103.

Changhao Shi, Chester Holtz, Gal Mishne, Online adversarial purification based on self-supervision, arXiv:2101.09387.

Dongwon Park, Daeyeal Lee, Ilgweon Kang, Chester Holtz, Sicun Gao, Bill Lin, Chung-Kuan Cheng, Grid-based framework for routability analysis and diagnosis with conditional design rules, ieee.

Chung-Kuan Cheng, Chia-Tung Ho, Chester Holtz, Net separation-oriented printed circuit board placement via margin maximization, arXiv:2210.14259.

Chester Holtz, Changhao Shi, Gal Mishne, Provable Robustness by Geometric Regularization of ReLU Networks, preprint.