Publications [#386391] of Xiuyuan Cheng
Papers Published
- Repasky, M; Cheng, X; Xie, Y. "STAGE-REGULARIZED NEURAL STEIN CRITICS FOR TESTING GOODNESS-OF-FIT OF GENERATIVE MODELS." ICASSP IEEE International Conference on Acoustics Speech and Signal Processing Proceedings (January, 2024): 7255-7259. [doi]
(last updated on 2026/01/19)Abstract:
Learning to differentiate model distributions from observed data is a fundamental problem in statistics and machine learning, and high-dimensional data remains a challenging setting for such problems. Metrics that quantify the disparity in probability distributions, such as the Stein discrepancy, play an important role in high-dimensional statistical testing. This paper presents a method based on neural network Stein critics to distinguish between data sampled from an unknown probability distribution and a nominal model distribution with a novel staging of the weight of regularization. The benefit of using staged L2 regularization in training such critics is demonstrated on evaluating generative models of image data.