Learning with Partially Labeled and Interdependent Data
This book develops two key machine learning principles: the semi-supervised paradigm and learning with interdependent data. It reveals new applications, primarily web related, that transgress the classical machine learning framework through learning with
- PDF / 1,835,411 Bytes
- 113 Pages / 439.44 x 666.24 pts Page_size
- 12 Downloads / 236 Views
Massih-Reza Amini • Nicolas Usunier
Learning with Partially Labeled and Interdependent Data
2123
Massih-Reza Amini Laboratoire d’Informatique de Grenoble Université Joseph Fourier Grenoble France
ISBN 978-3-319-15725-2 DOI 10.1007/978-3-319-15726-9
Nicolas Usunier Université Technologique de Compiègne Compiègne France
ISBN 978-3-319-15726-9 (e-Book)
Library of Congress Control Number: 2015933645 Springer Cham Heidelberg New York Dordrecht London © Springer International Publishing Switzerland 2015 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained herein or for any errors or omissions that may have been made. Printed on acid-free paper Springer is part of Springer Science+Business Media (www.springer.com)
All that is necessary, to reduce the whole of nature to laws similar to those which Newton discovered with the aid of the calculus, is to have a sufficient number of observations and a mathematics that is complex enough. Marquis Nicolas de Condorcet 1743–1794
Acknowledgements
The authors wish to thank the following those who made valuable suggestions or who have otherwise contributed to the preparation of the manuscript: Marianne Clausel, Cyril Goutte, François Laviolette, Clément Calauzènes, Patrick Gallinari and Guillaume Wisniewski.
vii
Contents
1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 New Learning Frameworks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1 1 2
2
Introduction to Learning Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Empirical Risk Minimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1 Assumption and Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.2 The Statement of the ERM Principle . . . . . . . . . . . . . . . . . . . . 2.2 Consistency of the ERM Principle . . . . . . . . . . . . . . . . . . . . . . .
Data Loading...
