Alert button

Kernel Change-point Detection with Auxiliary Deep Generative Models

Add code
Bookmark button
Alert button
Jan 18, 2019
Wei-Cheng Chang, Chun-Liang Li, Yiming Yang, Barnabás Póczos

Figure 1 for Kernel Change-point Detection with Auxiliary Deep Generative Models
Figure 2 for Kernel Change-point Detection with Auxiliary Deep Generative Models
Figure 3 for Kernel Change-point Detection with Auxiliary Deep Generative Models
Figure 4 for Kernel Change-point Detection with Auxiliary Deep Generative Models

Share this with someone who'll enjoy it:

Twitter IconFacebook IconLinkedin IconWhatsapp IconMessenger Icon

Detecting the emergence of abrupt property changes in time series is a challenging problem. Kernel two-sample test has been studied for this task which makes fewer assumptions on the distributions than traditional parametric approaches. However, selecting kernels is non-trivial in practice. Although kernel selection for two-sample test has been studied, the insufficient samples in change point detection problem hinder the success of those developed kernel selection algorithms. In this paper, we propose KL-CPD, a novel kernel learning framework for time series CPD that optimizes a lower bound of test power via an auxiliary generative model. With deep kernel parameterization, KL-CPD endows kernel two-sample test with the data-driven kernel to detect different types of change-points in real-world applications. The proposed approach significantly outperformed other state-of-the-art methods in our comparative evaluation of benchmark datasets and simulation studies.

* To appear in ICLR 2019  
View paper onarxiv iconopen_review iconOpenReview

Share this with someone who'll enjoy it:

Twitter IconFacebook IconLinkedin IconWhatsapp IconMessenger Icon