Stochastic automata operating in an unknown random can be considered to show learning behavior. Tsypkin environment have been proposed earlier as models of learning. These [GT1] has recently argued that seemingly diverse problems automata update their action probabilities in accordance with the inputs . . . received from the environment and can improve their own performance inpa t rec i o idenf...

An algorithm based on Newton’s Method is proposed for action selection in continuous stateand action-space reinforcement learning without a policy network or discretization. The proposed method is validated on two benchmark problems: Cart-Pole and double Cart-Pole on which the proposed method achieves comparable or improved performance with less parameters to tune and in less training episodes ...

This paper addresses the problem of knowledge transfer in lifelong reinforcement learning. It proposes an algorithm which learns policy constraints, i.e., rules that characterize action selection in entire families of reinforcement learning tasks. Once learned, policy constraints are used to bias learning in future, similar reinforcement learning tasks. The appropriateness of the algorithm is d...

Decision theory addresses the task of choosing an action; it provides robust decision-making criteria that support decision-making under conditions of uncertainty or risk. Decision theory has been applied to produce reinforcement learning algorithms that manage uncertainty in state-transitions. However, performance when there is uncertainty regarding the selection of future actions must also be...

Good representations can help RL agents perform concise modeling of their surroundings, and thus support effective decision-making in complex environments. Previous methods learn good by imposing extra constraints on dynamics. However, the causal perspective, causation between action its effect is not fully considered those methods, which leads to ignorance underlying relations among effects tr...

Multiaction learning automata which update their action probabilities on the basis of the responses they get from an environment are considered in this paper. The automata update the probabilities according to whether the environment responds with a reward or a penalty. Learning automata are said to possess ergodicity of the mean if the mean action probability is the state probability (or uncon...

We address the conflict between identification and control or alternatively, the conflict between exploration and exploitation, within the framework of reinforcement learning. Qlearning has recently become a popular offpolicy reinforcement learning method. The conflict between exploration and exploitation slows down Q-learning algorithms; their performance does not scale up and degrades rapidly...

Reinforcement learning in real-world domains suffers from three curses of dimensionality: explosions in state and action spaces, and high stochasticity. We present approaches that mitigate each of these curses. To handle the state-space explosion, we introduce “tabular linear functions” that generalize tile-coding and linear value functions. Action space complexity is reduced by replacing compl...

We report on an investigation of reinforcement learning techniques for the learning of coordination in cooperative multiagent systems. Specifically, we focus on a novel action selection strategy for Q-learning (Watkins 1989). The new technique is applicable to scenarios where mutual observation of actions is not possible. To date, reinforcement learning approaches for such independent agents di...

Learning algorithms for feedforward connectionist systems in a reinforcement learning environment are developed and analyzed in this paper. The connectionist system is made of units of groups of learning automata. The learning algorithm used is the LR-I and the asymptotic behavior of this algorithm is approximated by an Ordinary Differential Equation (ODE) for low values of the learning paramet...