COLT
2001
Springer
13 years 9 months ago
2001
Springer
Abstract. We develop three new techniques to build on the recent advances in online learning with kernels. First, we show that an exponential speed-up in prediction time per trial ...
COLT
2001
Springer
13 years 9 months ago
2001
Springer
In this paper we study a paradigm to generalize online classification algorithms for binary classification problems to multiclass problems. The particular hypotheses we investig...
COLT
2001
Springer
13 years 9 months ago
2001
Springer
In this paper we derive convergence rates for Q-learning. We show an interesting relationship between the convergence rate and the learning rate used in Q-learning. For a polynomi...
COLT
2001
Springer
13 years 9 months ago
2001
Springer
The Kushilevitz-Mansour (KM) algorithm is an algorithm that finds all the “large” Fourier coefficients of a Boolean function. It is the main tool for learning decision trees ...
COLT
2001
Springer
13 years 9 months ago
2001
Springer
We present results concerning the learning of Monotone DNF (MDNF) from Incomplete Membership Queries and Equivalence Queries. Our main result is a new algorithm that allows effici...
COLT
2001
Springer
13 years 9 months ago
2001
Springer
We prove strong noise-tolerance properties of a potential-based boosting algorithm, similar to MadaBoost (Domingo and Watanabe, 2000) and SmoothBoost (Servedio, 2003). Our analysi...
COLT
2001
Springer
13 years 9 months ago
2001
Springer
The notion of embedding a class of dichotomies in a class of linear half spaces is central to the support vector machines paradigm. We examine the question of determining the mini...
COLT
2001
Springer
13 years 9 months ago
2001
Springer
We investigate the use of certain data-dependent estimates of the complexity of a function class, called Rademacher and Gaussian complexities. In a decision theoretic setting, we ...
COLT
2001
Springer
13 years 9 months ago
2001
Springer
In this paper, we examine on-line learning problems in which the target concept is allowed to change over time. In each trial a master algorithm receives predictions from a large ...