A Survey of Reinforcement Learning Algorithms for Dynamically Varying Environments

  • 2020-05-19 09:42:42
  • Sindhu Padakandla
  • 1


Reinforcement learning (RL) algorithms find applications in inventorycontrol, recommender systems, vehicular traffic management, cloud computing androbotics. The real-world complications of many tasks arising in these domainsmakes them difficult to solve with the basic assumptions underlying classicalRL algorithms. RL agents in these applications often need to react and adapt tochanging operating conditions. A significant part of research on single-agentRL techniques focuses on developing algorithms when the underlying assumptionof stationary environment model is relaxed. This paper provides a survey of RLmethods developed for handling dynamically varying environment models. The goalof methods not limited by the stationarity assumption is to help autonomousagents adapt to varying operating conditions. This is possible either byminimizing the rewards lost during learning by RL agent or by finding asuitable policy for the RL agent which leads to efficient operation of theunderlying system. A representative collection of these algorithms is discussedin detail in this work along with their categorization and their relativemerits and demerits. Additionally we also review works which are tailored toapplication domains. Finally, we discuss future enhancements for this field.


Quick Read (beta)

loading the full paper ...