COLT 2000 Program

COLT 2000 Program

Wednesday June 28, 2000

Registration (8:30 - 9:00)
Opening Remarks(9:00 - 9:10)

Session 1 (9:10 - 10:30)

On the Convergence Rate of Good-Turing Estimators,
David McAllester and Robert E. Schapire

On the Efficiency of Noise-Tolerant PAC Algorithms Derived from Statistical Queries,
Jeffrey Jackson

Decision Tree Approximations of Boolean Functions,
Dinesh Mehta and Vijay Raghavan

Computable Shell Decomposition Bounds,
John Langford and David McAllester

Session 2 (11:00 - 12:00)

Continuous Drifting Games,
Yoav Freund and Manfred Opper

Language Learning From Texts: Degrees of Instrinsic Complexity and Their Characterizations,
Sanjay Jain, Efim Kinber and Rolf Wiehagen

Average-Case Complexity of Learning Polynomials,
Frank Stephan and Thomas Zeugmann

Session 3 (2:00 - 3:00)

Generalization Bounds for Decision Trees,
Yishay Mansour and David McAllester

The Role of Critical Sets in Vapnik-Chervonenkis Theory,
Nicolas Vayatis

Statistical Sufficiency for Classes in Empirical L_2 Spaces,
Shahar Mendelson and Naftali Tishby

Tutorial 1 (3:30 - 5:30)

Proving Relative Loss Bounds for On-Line Learning Algorithms by the Bregman Divergence,
Claudio Gentile and Manfred Warmuth

Reception and Barbecue (6:00 - 8:30)

COLT Business Meeting (8:30 - 9:30)

Thursday, June 29, 2000

Session 4 (9:00 - 10:20)

Relative Expected Instantaneous Loss Bounds,
Juergen Forster and Manfred Warmuth

The Minimax Strategy for Gaussian Density Estimation,
Eiji Takimoto and Manfred Warmuth

Adaptive and Self-Confident On-Line Learning Algorithms,
Peter Auer and Claudio Gentile

An Improved On-line Algorithm for Learning Linear Evaluation Functions,
Peter Auer

Session 5 (11:30 - 12:30)

On the Learnability and Design of Output Codes for Multiclass Problems,
Koby Crammer and Yoram Singer

Estimation and Approximation Bounds for Gradient-Based Reinforcement Learning,
Peter L. Bartlett and Jonathan Baxter

Bias-Variance Error Bounds for Temporal Difference Updates,
Michael Kearns and Satinder Singh

COLT Invited Talk (2:00 - 3:00)

Web Information Retrieval,
Monika Henzinger

Session 6 (3:30 - 4:50)

PAC Analogues of Perceptron and Winnow via Boosting the Margin,
Rocco A. Servedio

Logistic Regression, AdaBoost and Bregman Distances,
Michael Collins, Robert E. Schapire and Yoram Singer

Barrier Boosting,
G. Raetsch, M. Warmuth, S. Mika, T. Onoda, S. Lemm, and K.-R. Mueller

MadaBoost: A Modification of AdaBoost,
Carlos Domingo and Osamu Watanabe

COLT Impromptu Session (5:00 - 6:30)

COLT/ICML Joint Reception with the Anton Schwartz Jazz Band (7:00 - 10:00)

Friday, June 30, 2000

Session 7 (9:10 - 10:30)

Localized Boosting,
Ron Meir and Ran El-Yaniv and Shai Ben-David

Improving Algorithms for Boosting,
Javed A. Aslam

Leveraging for Regression,
Nigel Duffy and David Helmbold

Boosting Using Branching Programs,
Yishay Mansour and David McAllester

Session 8 (11:00 - 12:00)

The Precision of Query Points as a Resource for Learning Convex Polytopes with Membership Queries,
Paul Goldberg and Stephen Kwek

Improved Algorithms for Theory Revision with Queries,
Judy Goldsmith, Robert H. Sloan, Balazs Szorenyi, and Gyorgy Turan

Abstract Combinatorial Characterizations of Exact Learning via Queries,
Jose Luis Balcazar, Jorge Castro, and David Guijarro

Tutorial 2 (2:00 - 3:00)

On the Computational Complexity of Learning with Neural Networks,
Shai Ben-David

Tutorial 3, Joint with UAI (3:30 - 5:30)

An Introduction to Support Vector Machines and Other Kernel-based Learning Methods,
John Shawe-Taylor and Nello Cristianini

Joint COLT/ICML/UAI Poster Session (7:00 - 10:00 )

Saturday, July 1, 2000

Session 9 (9:10 - 10:10)

The Complexity of Densest Region Detection,
Shai Ben-David, Nadav Eiron, and Hans Simon

On the Difficulty of Approximately Maximizing Agreements,
Shai Ben-David, Nadav Eiron, and Philip M. Long

Hardness Results for General Two-Layer Neural Networks,
Christian Kuhlmann

Session 10 (10:40 - 12:00)

Model Selection and Error Estimation,
Peter L. Bartlett, Stephane Boucheron, and Gabor Lugosi

Generalisation Error Bounds for Sparse Linear Classifiers,
Thore Graepel, Ralf Herbrich, and John Shawe-Taylor

Sparsity vs. Large Margins for Linear Classifiers,
Ralf Herbrich, Thore Graepel, and John Shawe-Taylor

Entropy Numbers of Linear Function Classes,
Robert C. Williamson, Alex J. Smola, and Bernhard Scholkopf

COLT Conference Ends (12:00)

ICML Poster Session II is on July 1 from 7:00-10:00pm.

COLT attendees are welcome to attend. If you are staying on campus it's included in your dining pacakge. Otherwise a ticket must be purchased.