BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Pizza &amp\; AI February 19 - Microsoft Research/University of Cam
 bridge
DTSTART:20190222T173000Z
DTEND:20190222T190000Z
UID:TALK120436@talks.cam.ac.uk
CONTACT:Microsoft Research Cambridge Talks Admins
DESCRIPTION:*Speaker 1* - Gregor Simm\n*Title* - Exploring Chemical Reacti
 on Networks with Gaussian Processes\n*Abstract* - For the theoretical unde
 rstanding of chemical systems\, many costly quantum chemical calculations 
 have to be performed. Therefore\, approximate methods are often employed w
 hich suffer from low accuracy. We address this issue by applying Gaussian 
 processes to learn the error of these approximate methods. With our approa
 ch\, a fast and error-controlled exploration of chemical reaction networks
  is now in reach\n\n*Speaker 2* - Kamil Ciosek\n*Title* - Reinforcement Le
 arning with Continuous Actions: Fourier Policy Gradients and Expected Poli
 cy Gradients\n*Abstract* - The talk motivates the use of RL for continuous
  control tasks and gives a brief overview of actor-critic methods applied 
 in this setting. I describe recent advances in the field and discuss how t
 hese methods are derived. I discuss the advantages and disadvantages of es
 timators used in policy gradients and a framework for deriving them\, know
 n as Fourier Policy Gradients. I also discuss approaches to exploration us
 ed in this setting\, particularly a scheme arising from the Expected Polic
 y Gradients framework that uses the curvature of the critic to drive explo
 ration. Finally\, I give an example of how RL with continuous actions is u
 sed at Microsoft Research
LOCATION:Auditorium\, Microsoft Research Ltd\, 21 Station Road\, Cambridge
 \, CB1 2FB
END:VEVENT
END:VCALENDAR