02/02/2021

Variance Penalized On-Policy and Off-Policy Actor-Critic

Arushi Jain, Gandharv Patil, Ayush Jain, Khimya Khetarpal, Doina Precup

Keywords:

Abstract Paper Similar Papers

Abstract: Reinforcement learning algorithms are typically geared towards optimizing the expected return of an agent. However, in many practical applications, low variance in the return is desired to ensure the reliability of an algorithm. In this paper, we propose on-policy and off-policy actor-critic algorithms that optimize a performance criterion involving both mean and variance in the return. Previous work uses the second moment of return to estimate the variance indirectly. Instead, we use a much simpler recently proposed direct variance estimator which updates the estimates incrementally using temporal difference methods. Using the variance-penalized criterion, we guarantee the convergence of our algorithm to locally optimal policies for finite state action Markov decision processes. We demonstrate the utility of our algorithm in tabular and continuous MuJoCo domains. Our approach not only performs on par with actor-critic and prior variance-penalization baselines in terms of expected return, but also generates trajectories which have lower variance in the return.

The video of this talk cannot be embedded. You can watch it here:
https://slideslive.com/38949278
(Link will open in new window)
 0
 0
 0
 0
  Share    
This is an embedded video. Talk and the respective paper are published at AAAI 2021 virtual conference. If you are one of the authors of the paper and want to manage your upload, see the question "My papertalk has been externally embedded..." in the FAQ section.

Comments

Post Comment
no comments yet
code of conduct: tbd Characters remaining: 140

Similar Papers

 1
 0
 0
 0
 13:50 
 0
 0
 0
 0
 13:28 
 0
 0
 0
 0
 4:23 
 0
 0
 0
 0
 14:22 
 0
 0
 0
 0
 3:12 
 0
 0
 0
 0
 13:45 
 0
 0
 0
 0
 10:21 
 0
 0
 0
 0
 3:26 
 0
 0
 0
 0
 5:43 
 0
 0
 0
 0
 3:22 
 0
 0
 0
 0
 20:01 
 0
 0
 0
 0
 16:14 
 0
 0
 0
 0
 5:13 
 0
 0
 0
 0
 14:41 
 0
 0
 0
 0
 12:35 
 0
 0
 0
 0
 21:51 
 0
 0
 0
 0
 14:45 
 0
 0
 0
 0
 14:43 
 0
 0
 0
 0
 14:10 
 0
 0
 0
 0
 3:24 
 0
 0
 0
 0
 3:07 
 0
 0
 0
 0
 5:09 
 0
 0
 0
 0
 3:12 
 0
 0
 0
 0
 3:11 
 0
 0
 0
 0
 3:16 
 0
 0
 0
 0
 3:05 
 0
 0
 0
 0
 15:14 
 1
 0
 0
 0
 13:54 
 0
 0
 0
 0
 4:29 
 0
 0
 0
 0
 2:30 
 0
 0
 0
 0
 3:20 
 0
 0
 0
 0
 8:24 
 0
 0
 0
 0
 3:34 
 0
 0
 0
 0
 13:11 
 0
 0
 0
 0
 14:10 
 0
 0
 0
 0
 5:17 
 0
 0
 0
 0
 3:26 
 0
 0
 0
 0
 5:18 
 0
 0
 0
 0
 15:37 
 0
 0
 0
 0
 5:25