Reinforcement learning, one of the most active research areas in artificial intelligence, is a computational approach to learning whereby an agent tries to maximize the total amount of reward it receives when interacting with a complex, uncertain environment. In Reinforcement Learning, Richard Sutton and Andrew Barto provide a clear and simple account of the key ideas and algorithms of reinforcement learning. Their discussion ranges from the history of the field's intellectual foundations to the most recent developments and applications. The only necessary mathematical background is familiarity with elementary concepts of probability. The book is divided into three parts. Part I defines the reinforcement learning problem in terms of Markov decision processes. Part II provides basic solution methods: dynamic programming, Monte Carlo methods, and temporal-difference learning. Part III presents a unified view of the solution methods and incorporates artificial neural networks, eligibility traces, and planning; the two final chapters present case studies and consider the future of reinforcement learning.

/pdf/reinforcement-learning-an-introduction-rzxgej9p17.pdf

Reinforcement Learning: An Introduction

New developments in the science of learning science of learning overview mind and brain how experts differ from novices how children learn learning and transfer the learning environment curriculum, instruction and commnity effective teaching - examples in history, mathematics and science teacher learning technology to support learning conclusions from new developments in the science of learning.

How people learn: Brain, mind, experience, and school.

The neural basis of drug craving: An incentive-sensitization theory of addiction

Bayesian methods have garnered huge interest in cognitive science as an approach to models of cognition and perception. On the other hand, Bayesian methods for data analysis have not yet made much headway in cognitive science against the institutionalized inertia of 20th century null hypothesis significance testing (NHST). Ironically, specific Bayesian models of cognition and perception may not long endure the ravages of empirical verification, but generic Bayesian methods for data analysis will eventually dominate. It is time that Bayesian data analysis became the norm for empirical methods in cognitive science. This article reviews a fatal flaw of NHST and introduces the reader to some benefits of Bayesian data analysis. The article presents illustrative examples of multiple comparisons in Bayesian analysis of variance and Bayesian approaches to statistical power. Copyright © 2010 John Wiley & Sons, Ltd. For further resources related to this article, please visit the WIREs website.

Bayesian data analysis.

How do people know as much as they do with as little information as they get? The problem takes many forms; learning vocabulary from text is an especially dramatic and convenient case for research. A new general theory of acquired similarity and knowledge representation, latent semantic analysis (LSA), is presented and used to successfully simulate such learning and several other psycholinguistic phenomena. By inducing global knowledge indirectly from local co-occurrence data in a large body of representative text, LSA acquired knowledge about the full vocabulary of English at a comparable rate to schoolchildren. LSA uses no prior linguistic or perceptual similarity knowledge; it is based solely on a general mathematical learning method that achieves powerful inductive effects by extracting the right number of dimensions (e.g., 300) to represent objects and contexts. Relations to other theories, phenomena, and problems are sketched.

/pdf/a-solution-to-plato-s-problem-the-latent-semantic-analysis-4jfk7ozfms.pdf

A Solution to Plato's Problem: The Latent Semantic Analysis Theory of Acquisition, Induction, and Representation of Knowledge.

From insects to humans, Charles Gallistel explores the sophisticated computations performed in these ubiquitous but neglected domains of animal learning. How do animals represent space, time, number and rate? From insects to humans, Charles Gallistel explores the sophisticated computations performed in these ubiquitous yet neglected domains of animal learning. He proposes new and imaginative hypotheses about brain and mental processes and provides original insights about animal behavior using a computational-representational framework that is an exciting alternative to traditional associative theories of learning. Gallistel argues compellingly that experimental psychologists should begin to view the phenomena of learning within a framework that utilizes as the proper unit of analysis the computation and storage of a quantity, rather than the formation of an association that has been the basis of traditional learning theory. His approach reveals the formal structure of the environmental relationships that animals master to time and orient their behavior. It clarifies what representations different animals can and cannot compute and the nature of the computations by which animals derive these representations.The author backs up this thesis with studies that encompass a vast range of animal learning: animal navigation (the use of dead reckoning and cognitive maps); the mechanisms of timekeeping in the nervous system; the registration and utilization of time of occurrence (circadian phase) in learned behavior; the learning and use of temporal intervals and of numerosity; the computation of rates of occurrence; modern findings and theories of classical conditioning. Gallistel surveys the experimental literature in zoology, biology, neuroscience, and psychology that bears on those aspects of their environment that animals represent and the computations they perform in constructing and utilizing those representations. He reveals the fundamental role these representations play in learning and memory, and the implications of these findings in the search for the cellular basis of memory. The Organization of Learning is included in the series Learning, Development, and Conceptual Change, edited by Lila Gleitman, Susan Carey, Elissa Newport, and Elizabeth Spelke. A Bradford Book

The organization of learning

1. Focus on the Preschooler 2. Training Studies Reconsidered 3. More Capacity Than Meets the Eye: Direct Evidence 4. Number Concepts in the Preschooler? 5. What Numerosities Can the Young Child Represent? 6. How Do Young Children Obtain Their Representations of Numerosity? 7. The Counting Model 8. The Development of the How-To-Count Principles 9. The Abstraction and Order-Irrelevance Counting Principles 10. Reasoning about Number 11. Formal Arithmetic and the Young Child's Understanding of Number 12. What Develops and How Conclusions References Index

The Child's Understanding of Number

The child's understanding of number

Preverbal and verbal counting and computation.

The authors draw together and develop previous timing models for a broad range of conditioning phenomena to reveal their common conceptual foundations: First, conditioning depends on the learning of the temporal intervals between events and the reciprocals of these intervals, the rates of event occurrence. Second, remembered intervals and rates translate into observed behavior through decision processes whose structure is adapted to noise in the decision variables. The noise and the uncertainties consequent on it have both subjective and objective origins. A third feature of these models is their timescale invariance, which the authors argue is a very important property evident in the available experimental data. This conceptual framework is similar to the psychophysical conceptual framework in which contemporary models of sensory processing are rooted. The authors contrast it with the associative conceptual framework.

Charles R. Gallistel

Papers

The organization of learning

The Child's Understanding of Number

The child's understanding of number

Preverbal and verbal counting and computation.

Time, rate, and conditioning.