Скачать книгу

on>

      

       Active Learning

       Synthesis Lectures on Artificial Intelligence and Machine Learning

      Editor

       Ronald J. Brachman, Yahoo! Research

       William W. Cohen, Carnegie Mellon University

       Thomas Dietterich, Oregon State University

      Active Learning

      Burr Settles

      2012

      Planning with Markov Decision Processes: An AI Perspective

      Mausam and Andrey Kolobov

      2012

      Computational Aspects of Cooperative Game Theory

      Georgios Chalkiadakis, Edith Elkind, and Michael Wooldridge

      2011

      Representations and Techniques for 3D Object Recognition and Scene Interpretation

      Derek Hoiem and Silvio Savarese

      2011

      A Short Introduction to Preferences: Between Artificial Intelligence and Social Choice

      Francesca Rossi, Kristen Brent Venable, and Toby Walsh

      2011

      Human Computation

      Edith Law and Luis von Ahn

      2011

      Trading Agents

      Michael P. Wellman

      2011

      Visual Object Recognition

      Kristen Grauman and Bastian Leibe

      2011

      Learning with Support Vector Machines

      Colin Campbell and Yiming Ying

      2011

      Algorithms for Reinforcement Learning

      Csaba Szepesvári

      2010

      Data Integration: The Relational Logic Approach

      Michael Genesereth

      2010

      Markov Logic: An Interface Layer for Artificial Intelligence

      Pedro Domingos and Daniel Lowd

      2009

      Introduction to Semi-Supervised Learning

      XiaojinZhu and Andrew B.Goldberg

      2009

      Action Programming Languages

      Michael Thielscher

      2008

      Representation Discovery using Harmonic Analysis

      Sridhar Mahadevan

      2008

      Essentials of Game Theory: A Concise Multidisciplinary Introduction

      Kevin Leyton-Brown and Yoav Shoham

      2008

      A Concise Introduction to Multiagent Systems and Distributed Artificial Intelligence

      Nikos Vlassis

      2007

      Intelligent Autonomous Robotics: A Robot Soccer Case Study

      Peter Stone

      2007

      Copyright © 2012 by Morgan & Claypool

      All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means—electronic, mechanical, photocopy, recording, or any other except for brief quotations in printed reviews, without the prior permission of the publisher.

      Active Learning

      Burr Settles

       www.morganclaypool.com

      ISBN: 9781608457250 paperback

      ISBN: 9781608457267 ebook

      DOI 10.2200/S00429ED1V01Y201207AIM018

      A Publication in the Morgan & Claypool Publishers series

       SYNTHESIS LECTURES ON ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING

      Lecture #18

      Series Editors: Ronald J. Brachman, Yahoo Research

      William W. Cohen, Carnegie Mellon University

      Thomas Dietterich, Oregon State University

      Series ISSN

      Synthesis Lectures on Artificial Intelligence and Machine Learning

      Print 1939-4608 Electronic 1939-4616

       Active Learning

      Burr Settles

      Carnegie Mellon University

       SYNTHESIS LECTURES ON ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING #18

       ABSTRACT

      The key idea behind active learning is that a machine learning algorithm can perform better with less training if it is allowed to choose the data from which it learns. An active learner may pose “queries,” usually in the form of unlabeled data instances to be labeled by an “oracle” (e.g., a human annotator) that already understands the nature of the problem. This sort of approach is well-motivated in many modern machine learning and data mining applications, where unlabeled data may be abundant or easy to come by, but training labels are difficult, time-consuming, or expensive to obtain.

      This book is a general introduction to active learning. It outlines several scenarios in which queries might be formulated, and details many query selection algorithms which have been organized into four broad categories, or “query selection frameworks.” We also touch on some of the theoretical foundations of active learning, and conclude with an overview of the strengths and weaknesses of these approaches in practice, including a summary of ongoing work to address these open challenges and opportunities.

       KEYWORDS

      active learning, expected error reduction, hierarchical sampling, optimal experimental design, query by committee, query by disagreement, query learning, uncertainty sampling, variance reduction

      Dedicated to my family and friends, who keep me asking questions.

       Contents

       Preface

       Acknowledgments

       1 Automating Inquiry

       1.1 A Thought Experiment

       1.2 Active Learning

       1.3 Scenarios for Active Learning

      

Скачать книгу