Powell / Ryzhov Optimal Learning

1. Auflage 2012
ISBN: 978-1-118-30982-7
Verlag: John Wiley & Sons
Format: PDF
Kopierschutz: Adobe DRM (»Systemvoraussetzungen)

E-Book, Englisch, 414 Seiten, E-Book

Reihe: Wiley Series in Probability and Statistics

Learn the science of collecting information to make effectivedecisions
Everyday decisions are made without the benefit of accurateinformation. Optimal Learning develops the needed principlesfor gathering information to make decisions, especially whencollecting information is time-consuming and expensive. Designedfor readers with an elementary background in probability andstatistics, the book presents effective and practical policiesillustrated in a wide range of applications, from energy, homelandsecurity, and transportation to engineering, health, andbusiness.
This book covers the fundamental dimensions of a learningproblem and presents a simple method for testing and comparingpolicies for learning. Special attention is given to the knowledgegradient policy and its use with a wide range of belief models,including lookup table and parametric and for online and offlineproblems. Three sections develop ideas with increasing levels ofsophistication:
* Fundamentals explores fundamental topics, includingadaptive learning, ranking and selection, the knowledge gradient,and bandit problems
* Extensions and Applications features coverage of linearbelief models, subset selection models, scalar functionoptimization, optimal bidding, and stopping problems
* Advanced Topics explores complex methods includingsimulation optimization, active learning in mathematicalprogramming, and optimal continuous measurements
Each chapter identifies a specific learning problem, presentsthe related, practical algorithms for implementation, and concludeswith numerous exercises. A related website features additionalapplications and downloadable software, including MATLAB and theOptimal Learning Calculator, a spreadsheet-based package thatprovides an introduc-tion to learning and a variety ofpolicies for learning.

Autoren/Hrsg.

Powell, Warren B.

Ryzhov, Ilya O.

Über Autor(innen)

WARREN B. POWELL, PhD, is Professor of Operations Research andFinancial Engineering at Princeton University, where he is founderand Director of CASTLE Laboratory, a research unit that works withindustrial partners to test new ideas found in operations research.The recipient of the 2004 INFORMS Fellow Award, Dr. Powell is theauthor of Approximate Dynamic Programming: Solving the Curses ofDimensionality, Second Edition (Wiley).
ILYA O. RYZHOV, PhD, is Assistant Professor in the Department ofDecision, Operations, and Information Technologies at the Robert H.Smith School of Business at the University of Maryland. He has madefundamental contributions to bridge the fields of ranking andselection with multiarmed bandits and optimal learning withmathematical programming.