We consider the policy search approach to reinforcement learning. We show that if a "baseline distribution" is given (indicating roughly how often we expect a good policy to visit each state), then we can derive a policy search algorithm that terminates in a finite number of steps, and for which we can provide non-trivial performance guarantees. We also demonstrate this algorithm on several grid-world POMDPs, a planar biped walking robot, and a double-pole balancing problem.

Significant plasticity in sensory cortical representations can be driven in mature animals either by behavioural tasks that pair sensory stimuli with reinforcement, or by electrophysiological experiments that pair sensory input with direct stimulation of neuromodulatory nuclei, but usually not by sensory stimuli presented alone. Biologically motivated theories of representational learning, however, have tended to focus on unsupervised mechanisms, which may play a significant role on evolutionary or devel- opmental timescales, but which neglect this essential role of reinforce- ment in adult plasticity. By contrast, theoretical reinforcement learning has generally dealt with the acquisition of optimal policies for action in an uncertain world, rather than with the concurrent shaping of sensory representations. This paper develops a framework for representational learning which builds on the relative success of unsupervised generative- modelling accounts of cortical encodings to incorporate the effects of reinforcement in a biologically plausible way.

Successful application of reinforcement learning algorithms often involves considerable hand-crafting of the necessary non-linear features to reduce the complexity of the value functions and hence to promote convergence of the algorithm. In contrast, the human brain readily and autonomously finds the complex features when provided with sufficient training. Recent work in machine learning and neurophysiology has demonstrated the role of the basal ganglia and the frontal cortex in mammalian reinforcement learning. This paper develops and explores new learning algorithms that provides potential new approaches to the feature construction problem. The algorithms are compared and evaluated on the Acrobot task.

We introduce a new algorithm based on linear programming that approximates the differential value function of an average-cost Markov decision process via a linear combination of pre-selected basis functions. The algorithm carries out a form of cost shaping and minimizes a version of Bellman error. We establish an error bound that scales gracefully with the number of states without imposing the (strong) Lyapunov condition required by its counter- part in [6]. We propose a path-following method that automates selection of important algorithm parameters which represent coun- terparts to the "state-relevance weights" studied in [6]. Over the past few years, there has been a growing interest in linear programming (LP) approaches to approximate dynamic programming (DP).

Agents learning to act in a partially observable domain may need to overcome the problem of perceptual aliasing i.e., different states that appear similar but require different responses. This problem is exacer- bated when the agent's sensors are noisy, i.e., sensors may produce dif- ferent observations in the same state. We show that many well-known reinforcement learning methods designed to deal with perceptual alias- ing, such as Utile Suffix Memory, finite size history windows, eligibility traces, and memory bits, do not handle noisy sensors well. We suggest a new algorithm, Noisy Utile Suffix Memory (NUSM), based on USM, that uses a weighted classification of observed trajectories. We compare NUSM to the above methods and show it to be more robust to noise.

We introduce a generalization of temporal-difference (TD) learning to networks of interrelated predictions. Rather than relating a single pre- diction to itself at a later time, as in conventional TD methods, a TD network relates each prediction in a set of predictions to other predic- tions in the set at a later time. TD networks can represent and apply TD learning to a much wider class of predictions than has previously been possible. Using a random-walk example, we show that these networks can be used to learn to predict by a fixed interval, which is not possi- ble with conventional TD methods. Secondly, we show that if the inter- predictive relationships are made conditional on action, then the usual learning-efficiency advantage of TD methods over Monte Carlo (super- vised learning) methods becomes particularly pronounced.

Psychologists call behavior intrinsically motivated when it is engaged in for its own sake rather than as a step toward solving a specific problem of clear practical value. But what we learn during intrinsically motivated behavior is essential for our development as competent autonomous en- tities able to efficiently solve a wide range of practical problems as they arise. In this paper we present initial results from a computational study of intrinsically motivated reinforcement learning aimed at allowing arti- ficial agents to construct and extend hierarchies of reusable skills that are needed for competent autonomy. Psychologists distinguish between extrinsic motivation, which means being moved to do something because of some specific rewarding outcome, and intrinsic motivation, which refers to being moved to do something because it is inherently enjoyable. Intrinsic motiva- tion leads organisms to engage in exploration, play, and other behavior driven by curiosity in the absence of explicit reward. These activities favor the development of broad com- petence rather than being directed to more externally-directed goals (e.g., ref. [14]). In contrast, machine learning algorithms are typically applied to single problems and so do not cope flexibly with new problems as they arise over extended periods of time. Although the acquisition of competence may not be driven by specific problems, this com- petence is routinely enlisted to solve many different specific problems over the agent's lifetime.

This paper proposes an algorithm to convert a T -stage stochastic decision problem with a continuous state space to a sequence of supervised learning problems. The optimization problem associated with the trajectory tree and random trajectory methods of Kearns, Mansour, and Ng, 2000, is solved using the Gauss-Seidel method. The algorithm breaks a multistage reinforcement learning problem into a sequence of single-stage reinforcement learning subproblems, each of which is solved via an exact reduction to a weighted-classification problem that can be solved using off-the-self methods. Thus the algorithm converts a reinforcement learning problem into simpler supervised learning subproblems. It is shown that the method converges in a finite number of steps to a solution that cannot be further improved by componentwise optimization.

We introduce a new algorithm for off-policy temporal-difference learning with function approximation that has lower variance and requires less knowledge of the behavior policy than prior methods. We develop the notion of a recognizer, a filter on actions that distorts the behavior policy to produce a related target policy with low-variance importance-sampling corrections. We also consider target policies that are deviations from the state distribution of the behavior policy, such as potential temporally abstract options, which further reduces variance. This paper introduces recognizers and their potential advantages, then develops a full algorithm for linear function approximation and proves that its updates are in the same direction as on-policy TD updates, which implies asymptotic convergence. Even though our algorithm is based on importance sampling, we prove that it requires absolutely no knowledge of the behavior policy for the case of state-aggregation function approximators.

Probabilistic temporal planning attempts to find good policies for acting in domains with concurrent durative tasks, multiple uncertain outcomes, and limited resources. These domains are typically modelled as Markov decision problems and solved using dynamic programming methods. This paper demonstrates the application of reinforcement learning -- in the form of a policy-gradient method -- to these domains. Our emphasis is large domains that are infeasible for dynamic programming. Our ap- proach is to construct simple policies, or agents, for each planning task.