Hiroaki Sasaki - Robust time contrastive learning (RTCL)

Robust time contrastive learning (RTCL)

Robust time contrastive learning (RTCL) is a nonlinear ICA method based on the γ-divergence, and is very robust against outliers.

Python implementation of RTCL: RTCL.tgz

My python implementation is based an implementation of time contrastive learning (TCL) by Dr. Hiroshi Morioka. The original TCL implementation is available at https://sites.google.com/view/hiroshimorioka/.

References

Hiroaki Sasaki, Takashi Takenouchi, Ricardo Monti and Aapo Hyvärinen, "Robust Contrastive Learning and Nonlinear ICA in the Presence of Outliers", Conference on Uncertainty in Artificial Intelligence (UAI), Proceedings of Machine Learning Research, vol.124, pp.659-669, 2020.

Google Sites

Report abuse