(Enter summary)
Abstract: We derive a new bound on the error rate for decision trees. The bounds depends both on
the structure of the tree and the specific sample (not just the size of the sample). This bound
is tighter than traditional bounds for unbalanced trees and justifies "compositional" algorithms
for constructing decision trees.
1 Introduction
The problem of over-fitting is central to both the theory and practice of machine learning. Intuitively,
one over-fits by using too many parameters in the concept, e.g,... (Update)
Context of citations to this paper: More
.... ; k m ] In either case we have ddpee 2 f 1 m ; m m g and if ddpee = k m then p 2 1 A derivation of this formula can be found in [8] or [9] To see the need for the last term consider the case where p = 0. k Gamma1 m ; k m ] Now we define H( k m ) to be the set of h 2...
.... m ; k m ] In either case we have ddpee 2 f 1 m ; m m g and if ddpee = k m then p 2 1 A derivation of this formula can be found in [8] or [9] To see the need for the last term consider the case where p = 0. k 1 m ; k m ] Now we define H( k m ) to be the set of h 2...
Cited by: More
Near-Minimax Optimal Classification with Dyadic Classification.. - Scott, Nowak (2003)
(Correct)
Improving Supervised Learning by Feature Decomposition - Maimon, Rokach
(Correct)
Theory and Applications of Attribute Decomposition - Department (2001)
(Correct)
Similar documents (at the sentence level):
70.8%: Generalization Bounds for Decision Trees - Mansour, McAllester (2000)
(Correct)
Active bibliography (related documents): More All
0.2: Quantitatively Tight Sample Complexity Bounds - Langford (2002)
(Correct)
0.1: Journal of Machine Learning Research 3 (2002) 463-482.. - Risk Bounds And
(Correct)
0.1: Rademacher and Gaussian Complexities: Risk Bounds and.. - Bartlett, Mendelson
(Correct)
Similar documents based on text: More All
0.1: Why Averaging Classifiers Can Protect Against Overfitting - Freund, Mansour, Schapire (2001)
(Correct)
0.1: Broadcast in Radio Networks - Gaber, Mansour (1995)
(Correct)
0.1: Pessimistic Decision Tree Pruning Based on Tree Size - Mansour (1997)
(Correct)
Related documents from co-citation: More All
2: Probability and Statistics for Engineers and Scientists (context) - Walpole, Myers - 1978
2: Applications of Information Theory to Psychology (context) - Attneave - 1959
2: Rigorous learning curve bounds from statistical mechanics
- Haussler, Kearns et al. - 1996
BibTeX entry: (Update)
Yishay Mansour and David McAllester, "Generalization Bounds for Decision Trees", COLT-2000. http://citeseer.ist.psu.edu/article/mansour99generalization.html More
@inproceedings{ mansour00generalization,
author = "Yishay Mansour and David Mc{A}llester",
title = "Generalization Bounds for Decision Trees",
booktitle = "Proc. 13th Annu. Conference on Comput. Learning Theory",
publisher = "Morgan Kaufmann, San Francisco",
pages = "69--80",
year = "2000",
url = "citeseer.ist.psu.edu/article/mansour99generalization.html" }
Citations (may not include all citations):
947
Statistical Learning Theory (context) - Vapnik - 1998
348
Estimation of Dependences Based on Empirical Data (context) - Vapnik - 1982
88
Neural Network Learning Theoretical Foundations (context) - Anthony, Bartlett - 1999
46
Some PAC-Bayesian theorems
- McAllester - 1998
24
Self bounding learning algorithms
- Freund - 1998
11
bottom-up decision tree pruning algorithm with near-optimal .. (context) - Kearns, Mansour et al. - 1998
9
Microchoice bounds and selfbounding learning algorithms
- Langford, Blum - 1999
7
Generalization in decision trees and DNF: Does size matter
- Golea, Bartlett et al. - 1997
Documents on the same site (http://www.research.att.com/~dmac/index.html): More
Automatic Recognition of Tractability in Inference Relations - McAllester (1993)
(Correct)
Natural Language Based Inference Procedures applied to.. - Givan, McAllester.. (1991)
(Correct)
Walther Recursion - McAllester, Arkoudas (1996)
(Correct)
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC