Downloads: 110 | Views: 172
Survey Paper | Computer Science & Engineering | India | Volume 4 Issue 5, May 2015
Adaptive Reinforcement Learning Method for Sequential Decision Task: A Review
Pramod Patil [5] | Ankur Verma
Abstract: There are many dynamic situations in which sequential actions come with circumstances favorable. These consequences of actions can include at a multitude of times after the action is taken, and it shall be concern with the strategies for specify action on the basis of both their short term and long term consequences. A proposed model based approach which requires constructing the model of state transaction and payoff probabilities. Task of such kind can be termed as a dynamical system whose behavior changes over time under the impact of a decision maker-s action. This modeling of the behavior of the system is greatly simplified by the concept of state. Decision policy associates on action with each system states. There is a great practical importance of adaptive method, if this adaptive method can make improvement in decision policy sufficiently rapidly may be less. It proposes methods for estimating optimal policy in the absence of a complete model of the decision tasks which are known as adaptive or decision model.
Keywords: Reinforcement Learning, Decision policy, state-action function, Q-Learning, Temporal Difference Learning
Edition: Volume 4 Issue 5, May 2015,
Pages: 2365 - 2368