We propose a new #ReinforcementLearning method for LQG control systems (hidden states) with surprising logarithmic regret. Our method is episodic + online learning: episodic state estimation with online convex optimization. @SahinLale @kazizzad https://arxiv.org/pdf/2003.11227.pdf …
-
-
Show this thread
-
Second part of my talk will cover control under non-linear dynamics. Distributionally robust learning methods with uncertainty bounds: safe exploration, planning and control.
@anqi_liu33@GuanyaShi @yash92006@yisongyue https://arxiv.org/pdf/2005.04374.pdf … https://arxiv.org/abs/1906.05819v1 …Show this thread -
Link to talk video http://tiny.cc/tmlseminaranima and slides http://tensorlab.cms.caltech.edu/users/anima/slides/Harvard-May2020.pptx … Thank you
@boazbaraktcs for organizing this Video has an active chat transcript. Thanks@SahinLale@anqi_liu33 for answering chat questions. Thanks@CsabaSzepesvari for asking so many hard questions ;)Show this thread
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.