当前位置:首页>学术报告
张驰浩 博士:Robust Bayesian Matrix Decomposition with Mixture of Gaussian Noise
浏览次数:

 
Title:
Robust Bayesian Matrix Decomposition with Mixture of Gaussian Noise
Speaker:
张驰浩 博士,日本东京大学
Inviter: 张世华 研究员
Time & Venue:

2021.10.27 8:40 N625

Abstract:

Matrix decomposition is a popular and fundamental approach in machine learning. The classical matrix decomposition methods with Frobenius norm loss is only optimal for Gaussian noise and thus suffer from the sensitivity to outliers and non-Gaussian noise. To address these limitations, the proposed methods can be divided into two categories. One type of approach is to replace the Frobenius norm loss with robust loss functions. The other type of approach is to impose the Bayesian priors to reduce the risk of overfitting. This paper combines these two approaches. Specifically, we model the noise by a mixture of Gaussian distribution, enabling the model to approximate a wide range of noise distributions. Meanwhile, we put a Laplace prior on the basis matrix to enforce the sparsity and a Dirichlet prior on the coefficient matrix to improve the interpretability. Extensive experiments in synthetic data and real-world data demonstrate that this method outperforms several competing ones. Ablation studies show that this method benefits from both the Bayesian priors and the Mixture of Gaussian noise loss, which confirms the necessity of combining the two schemes.

Affiliation:  

学术报告中国科学院数学与系统科学研究院应用数学研究所
地址 北京市海淀区中关村东路55号 思源楼6-7层 南楼5-6、8层 100190
?2000-2013 京ICP备05058656号