5.0 out of 5 stars
Review from the author of Approximate Dynamic Programming, December 15, 2007
If you believe in the axiom "less is more," this is an outstandingbook. This is the book that attracted me to the field of dynamicprogramming. The presentation is exceptionally clear, and gives anintroduction to the simple, elegant problems that makes the field soaddictive. It takes only a few afternoons to go through the entirebook. In fact, it was memories of this book that guided theintroduction to my own book on approximate dynamic programming (seechapter 2).
Once you have been drawn to the field with this book, you will wantto trade up to Puterman's much more thorough presentation in
Markov Decision Processes: Discrete Stochastic Dynamic Programming (Wiley Series in Probability and Statistics).But be forewarned - this elegant theory, which uses a "flatrepresentation" of states (where states are numbered 1, 2, ..., S),suffers from the well-known curse of dimensionality, limiting itspractical application. If your interests are drawn to real problems,you might consider my recent book
Approximate Dynamic Programming: Solving the Curses of Dimensionality (Wiley Series in Probability and Statistics),which puts far more emphasis on modeling and practical algorithms drawnfrom the field of approximate dynamic programming. Other importantreferences in this field are
Neuro-Dynamic Programming (Optimization and Neural Computation Series, 3), and
Reinforcement Learning: An Introduction (Adaptive Computation and Machine Learning).
Warren B. Powell
Professor
Princeton University
扫码加好友，拉您进群
全部版块
我的主页
收藏
