Искусственный интеллект
Информатика и вычислительная техника
  • формат pdf
  • размер 1.27 МБ
  • добавлен 26 октября 2011 г.
Mahadevan S. Learning Representation and Control in Markov Decision Processes: New Frontiers
Из серии Foundations and Trends in Machine Leaing издательства NOWPress, 2008, -163 pp.

This paper describes a novel machine leaing framework for solving sequential decision problems called Markov decision processes (MDPs) by iteratively computing low-dimensional representations and approximately optimal policies. A unified mathematical framework for leaing representation and optimal control in MDPs is presented based on a class of singular operators called Laplacians, whose matrix representations have nonpositive off-diagonal elements and zero row sums. Exact solutions of discounted and average-reward MDPs are expressed in terms of a generalized spectral inverse of the Laplacian called the Drazin inverse. A generic algorithm called representation policy iteration (RPI) is presented which interleaves computing low-dimensional representations and approximately optimal policies. Two approaches for dimensionality reduction of MDPs are described based on geometric and reward-sensitive regularization, whereby low-dimensional representations are formed by diagonalization or dilation of Laplacian operators. Model-based and model-free variants of the RPI algorithm are presented; they are also compared experimentally on discrete and continuous MDPs. Some directions for future work are finally outlined.

Introduction
Sequential Decision Problems
Laplacian Operators and MDPs
Approximating Markov Decision Processes
Dimensionality Reduction Principles in MDPs
Basis Construction: Diagonalization Methods
Basis Construction: Dilation Methods
Model-Based Representation Policy Iteration
Basis Construction in Continuous MDPs
Model-Free Representation Policy Iteration
Related Work and Future Challenges
Похожие разделы
Смотрите также

Alpaydin E. Introduction to Machine Learning

  • формат pdf
  • размер 2.87 МБ
  • добавлен 05 октября 2011 г.
Издательство MIT Press, 2010, -581 pp. Machine learning is programming computers to optimize a performance criterion using example data or past experience. We need learning in cases where we cannot directly write a computer program to solve a given problem, but need example data or experience. One case where learning is necessary is when human expertise does not exist, or when humans are unable to explain their expertise. Consider the recognitio...

Duda R.O., Hart P.E., Stork D.G. Pattern classification (2nd edition)

  • формат djvu
  • размер 6.4 МБ
  • добавлен 16 мая 2010 г.
2001, 738 pages. The ease with which we recognize a face, understand spoken words, read handwritten characters, identify our car keys in our pocket by feel, and decide whether an apple is ripe by its smell belies the astoundingly complex processes that underlie these acts of pattern recognition. Pattern recognition — the act of taking in raw data and taking an action based on the "category" of the pattern — has been crucial for our survival, and...

Er M.J., Zhou Y. (eds.) Theory and Novel Applications of Machine Learning

  • формат pdf
  • размер 6.75 МБ
  • добавлен 12 ноября 2011 г.
Издательство InTech, 2009, -386 pp. Even since computers were invented many decades ago, many researchers have been trying to understand how human beings learn and many interesting paradigms and approaches towards emulating human learning abilities have been proposed. The ability of learning is one of the central features of human intelligence, which makes it an important ingredient in both traditional Artificial Intelligence (AI) and emerging...

Fiedler K., Juslin P. (eds.) Information Sampling and Adaptive Cognition

  • формат pdf
  • размер 2.59 МБ
  • добавлен 30 января 2012 г.
Издательство Cambridge University Press, 2006, -498 pp. A sample is not only a concept from statistics that has penetrated common sense but also a metaphor that has inspired much research and theorizing in current psychology. The sampling approach emphasizes the selectivity and biases inherent in the samples of information input with which judges and decision makers are fed. Because environmental samples are rarely random, or representative of t...

Rasmussen C.E., Williams C.K.I. Gaussian Processes for Machine Learning

  • формат pdf
  • размер 3.86 МБ
  • добавлен 16 декабря 2011 г.
Издательство MIT Press, 2006, -266 pp. The book is primarily intended for graduate students and researchers in machine learning at departments of Computer Science, Statistics and Applied Mathematics. As prerequisites we require a good basic grounding in calculus, linear algebra and probability theory as would be obtained by graduates in numerate disciplines such as electrical engineering, physics and computer science. For preparation in calculus...

Setlak G., Markov K. Methods and Instruments of Artificial Intelligence

Статья
  • формат pdf
  • размер 4.43 МБ
  • добавлен 25 октября 2010 г.
ITHEA ® Материалы конференции Rzeszow, Poland – Sofia, Bulgaria, 2010 ISBN 978-954-16-0049-8 This book maintains articles on actual problems of research and application of information technologies, especially the new approaches, models, algorithms and methods for Hybrid Intelligent Systems; Intelligent Agents and Multi-Agent Systems; Software Engineering and Development; Knowledge Representation and Management; Intelligent Robots; Intelligent...

Wang C., Hill D.J. Deterministic Learning Theory for Identification, Recognition and Control

  • формат pdf
  • размер 10.94 МБ
  • добавлен 29 ноября 2011 г.
Издательство CRC Press, 2010, -218 pp. The problem of learning in dynamic environments is important and challenging. In the 1960s, learning from control of dynamical systems was studied extensively. At that time, learning was similar in meaning to other terms such as adaptation and self-organizing. Since the 1970s, learning theory has become a research discipline in the context of machine learning, and more recently as computational or statistic...

Weber C., Elshaw M., Mayer N.M. (eds.) Reinforcement Learning. Theory and Applications

  • формат pdf
  • размер 11.72 МБ
  • добавлен 12 ноября 2011 г.
Издательство InTech, 2011, -434 pp. Brains rule the world, and brain-like computation is increasingly used in computers and electronic devices. Brain-like computation is about processing and interpreting data or directly putting forward and performing actions. Learning is a very important aspect. This book is on reinforcement learning which involves performing actions to achieve a goal. Two other learning paradigms exist. Supervised learning ha...

Wilamowski B.M., Irwin J.D. Intelligent Systems (The Industrial Electronics Handbook, Second Edition)

  • формат pdf
  • размер 25.49 МБ
  • добавлен 22 августа 2011 г.
CRC Press, 2011. - 568 p. The field of industrial electronics covers a plethora of problems which must be solved in industrial practice. Electronic systems control many processes that begin with the control of relatively simple devices like electric motors, through more complicated devices such as robots, to the control of entire fabrication processes. An industrial electronics engineer works with many physical phenomena as well as the sensors...