This paper surveys the field of reinforcement learning from a computer-science perspective. It is written to be accessible to researchers familiar with machine learning. Both the historical basis of the field and a broad selection of current work are summarized. Reinforcement learning is the problem faced by an agent that learns behavior through trial-and-error interactions with a dynamic environment. The work described here has a resemblance to work in psychology, but differs considerably in the details and in the use of the word "reinforcement." The paper discusses central issues of reinforcement learning, including trading off exploration and exploitation, establishing the foundations of the field via Markov decision theory, learning from delayed reinforcement, constructing empirical models to accelerate learning, making use of generalization and hierarchy, and coping with hidden state. It concludes with a survey of some implemented systems and an assessment of the practical utility of current methods for reinforcement learning.

� How should a robot decide what to do? � It should plan for each move (Planning) � It should plan for all moves and compile its results into a set of rapid reactions (Reactive Systems) � It should Learn a set of reactions by trial-anderror

Abstract not found

We present a new algorithm, Prioritized Sweeping, for e cient prediction and control of stochas-tic Markov systems. Incremental learning methods such asTemporal Di erencing and Q-learning have fast real time performance. Classical methods are slower, but more accurate, because they make full use of the observations. Prioritized Sweeping aims for the best of both worlds. It uses all previous experiences both to prioritize important dynamic programming sweeps and to guide the exploration of state-space. We compare Prioritized Sweeping with other reinforcement learning schemes for a number of di erent stochastic optimal control prob-lems. It successfully solves large state-space real time problems with which other methods have di culty. 1 1

Abstract. To date, reinforcement learning has mostly been studied solving simple learning tasks. Reinforcement learning methods that have been studied so far typically converge slowly. The purpose of this work is thus two-fold: 1) to investigate the utility of reinforcement learning in solving much more complicated learning tasks than previously studied, and 2) to investigate methods that will speed up reinforcement learning. This paper compares eight reinforcement learning frameworks: adaptive heuristic critic (AHC) learning due to Sutton, Q-learning due to Watkins, and three extensions to both basic methods for speeding up learning. The three extensions are experience replay, learning action models for planning, and teaching. The frameworks were investigated using connectionism as an approach to generalization. To evaluate the performance of different frame-works, a dynamic environment was used as a testbed. The enviromaaent is moderately complex and nondetermin-istic. This paper describes these frameworks and algorithms in detail and presents empirical evaluation of the frameworks.

By now it is widely accepted that learning a task from scratch, i.e., without any prior knowledge, is a daunting undertaking. Humans, however, rarely attempt to learn from scratch. They extract initial biases as well as strategies how to approach a learning problem from instructions and/or demonstrations of other humans. For learning control, this paper investigates how learning from demonstration can be applied in the context of reinforcement learning. We consider priming the Q-function, the value function, the policy, and the model of the task dynamics as possible areas where demonstrations can speed up learning. In general nonlinear learning problems, only model-based reinforcement learning shows significant speed-up after a demonstration, while in the special case of linear quadratic regulator (LQR) problems, all methods profit from the demonstration. In an implementation of pole balancing on a complex anthropomorphic robot arm, we demonstrate that, when facing the complexities of real signal processing, model-based reinforcement learning offers the most robustness for LQR problems. Using the suggested methods, the robot learns pole balancing in just a single trial after a 30 second long demonstration of the human instructor. 1.

Internal models of the environment have an important role to play in adaptive systems in general and are of particular importance for the supervised learning paradigm. In this paper we demonstrate that certain classical problems associated with the notion of the \teacher " in supervised learning can be solved by judicious use of learned internal models as components of the adaptive system. In particular, we show how supervised learning algorithms can be utilized in cases in which an unknown dynamical system intervenes between actions and desired outcomes. Our approach applies to any supervised learning algorithm that is capable of learning in multi-layer networks.

In this paper, we describe the partially observable Markov decision process (pomdp) approach to finding optimal or near-optimal control strategies for partially observable stochastic environments, given a complete model of the environment. The pomdp approach was originally developed in the operations research community and provides a formal basis for planning problems that have been of interest to the AI community. We found the existing algorithms for computing optimal control strategies to be highly computationally inefficient and have developed a new algorithm that is empirically more efficient. We sketch this algorithm and present preliminary results on several small problems that illustrate important properties of the pomdp approach.

Abstract. Parti-game is a new algorithm for learning feasible trajectories to goal regions in high dimensional continuous state-spaces. In high dimensions it is essential that learning does not plan uniformly over a state-space. Parti-game maintains a decision-tree partitioning of state-space and applies techniques from game-theory and computational geometry to e ciently and adaptively concentrate high resolution only on critical areas. The currentversion of the algorithm is designed to nd feasible paths or trajectories to goal regions in high dimensional spaces. Future versions will be designed to nd a solution that optimizes a real-valued criterion. Many simulated problems have been tested, ranging from two-dimensional to nine-dimensional state-spaces, including mazes, path planning, non-linear dynamics, and planar snake robots in restricted spaces. In all cases, a good solution is found in less than ten trials and a few minutes.

We describe an algorithm which allows a behavior-based robot to learn on the basis of positive and negative feedback when to activate its behaviors. In accordance with the philosophy of behavior-based robots, the algorithm is completely distributed: each of the behaviors independently tries to find out (i) whether it is relevant (ie. whether it is at all correlated to positive feedback) and (ii) what the conditions are under which it becomes reliable (i.e. the conditions under which it maximizes the probability of receiving positive feedback and minimizes the probability of receiving negative feedback). The algorithm has been tested successfully on an autonomous 6-legged robot which had to learn how to coordinate its legs so as to walk forward. Situation of the Problem Since 1985, the MIT Mobile Robot group has advocated a radically different architecture for autonomous intelligent agents (Brooks, 1986). Instead of decomposing the architecture into functional modules, such as percept...