Evolutionary Computation for Reinforcement Learning
نویسنده
چکیده
Algorithms for evolutionary computation, which simulate the process of natural selection to solve optimization problems, are an effective tool for discovering high-performing reinforcement-learning policies. Because they can automatically find good representations, handle continuous action spaces, and cope with partial observability, evolutionary reinforcement-learning approaches have a strong empirical track record, sometimes significantly outperforming temporal-difference methods. This chapter surveys research on the application of evolutionary computation to reinforcement learning, overviewing methods for evolving neural-network topologies and weights, hybrid methods that also use temporal-difference methods, coevolutionary methods for multi-agent settings, generative and developmental systems, and methods for on-line evolutionary reinforcement learning.
منابع مشابه
Multicast Routing in Wireless Sensor Networks: A Distributed Reinforcement Learning Approach
Wireless Sensor Networks (WSNs) are consist of independent distributed sensors with storing, processing, sensing and communication capabilities to monitor physical or environmental conditions. There are number of challenges in WSNs because of limitation of battery power, communications, computation and storage space. In the recent years, computational intelligence approaches such as evolutionar...
متن کاملReinforcement Learning through Evolutionary Computation
This article characterizes the evolutionary algorithm approach to reinforcement learning in relation to the more standard, temporal diierence methods. We describe several research issues in reinforcement learning and discuss similarities and diierences in how they are addressed by the two methods. A short survey of evolutionary reinforcement learning systems and their successful applications is...
متن کاملA Proposal for Population-Based Reinforcement Learning
We propose novel ways of solving Reinforcement Learning tasks (that is, stochastic optimal control tasks) by hybridising Evolutionary Algorithms with methods based on value functions. We call our approach Population-Based Reinforcement Learning. The key idea, from Evolutionary Computation, is that parallel interacting search processes (in this case Reinforcement Learning or Dynamic Programming ...
متن کاملEvolutionary Function Approximation for Reinforcement Learning
Temporal difference methods are theoretically grounded and empirically effective methods for addressing reinforcement learning problems. In most real-world reinforcement learning tasks, TD methods require a function approximator to represent the value function. However, using function approximators requires manually making crucial representational decisions. This paper investigates evolutionary...
متن کاملIntegrating System Optimum and User Equilibrium in Traffic Assignment via Evolutionary Search and Multiagent Reinforcement Learning
Traffic assignment is fundamentally a tool for transportation planning. It allocates trips within the traffic network. However, modern uses of traffic assignment also include shorter time horizons and even real-time use (e.g., for route recommendation). In the latter case, it is interesting to recommend routes that are as close as possible to the system optimum. To compute an approximation of t...
متن کامل