DeepAI AI Chat
Log In Sign Up

Evolution Strategies as a Scalable Alternative to Reinforcement Learning

by   Tim Salimans, et al.

We explore the use of Evolution Strategies (ES), a class of black box optimization algorithms, as an alternative to popular MDP-based RL techniques such as Q-learning and Policy Gradients. Experiments on MuJoCo and Atari show that ES is a viable solution strategy that scales extremely well with the number of CPUs available: By using a novel communication strategy based on common random numbers, our ES implementation only needs to communicate scalars, making it possible to scale to over a thousand parallel workers. This allows us to solve 3D humanoid walking in 10 minutes and obtain competitive results on most Atari games after one hour of training. In addition, we highlight several advantages of ES as a black box optimization technique: it is invariant to action frequency and delayed rewards, tolerant of extremely long horizons, and does not need temporal discounting or value function approximation.


page 1

page 2

page 3

page 4


COCOpf: An Algorithm Portfolio Framework

Algorithm portfolios represent a strategy of composing multiple heuristi...

GPU Accelerated Exhaustive Search for Optimal Ensemble of Black-Box Optimization Algorithms

Black-box optimization is essential for tuning complex machine learning ...

Regret-Aware Black-Box Optimization with Natural Gradients, Trust-Regions and Entropy Control

Most successful stochastic black-box optimizers, such as CMA-ES, use ran...

Improving Sample Efficiency in Evolutionary RL Using Off-Policy Ranking

Evolution Strategy (ES) is a powerful black-box optimization technique b...

Black-box Optimizers vs Taste Shocks

We evaluate and extend the solution methods for models with binary and m...

Noise-Reuse in Online Evolution Strategies

Online evolution strategies have become an attractive alternative to aut...

Instance Weighted Incremental Evolution Strategies for Reinforcement Learning in Dynamic Environments

Evolution strategies (ES), as a family of black-box optimization algorit...

Code Repositories


Contains implementation of: Tim Salimans, Jonathan Ho, Xi Chen, and Ilya Sutskever. “Evolution Strategies as a Scalable Alternative to Reinforcement Learning”.

view repo


Project for Madhacks Fall 2017

view repo