Abstract

The areas of On-Line Algorithms and Machine Learning are both concerned with problems of making decisions about the present based only on knowledge of the past. Although these areas differ in terms of their emphasis and the problems typically studied, there are a collection of results in Computational Learning Theory that fit nicely into the "on-line algorithms" framework. This survey article discusses some of the results, models, and open problems from Computational Learning Theory that seem particularly interesting from the point of view of on-line algorithms. The emphasis in this article is on describing some of the simpler, more intuitive results, whose proofs can be given in their entirity. Pointers to the literature are given for more sophisticated versions of these algorithms.

Citations

1515	A theory of the learnable - Valiant - 1984
605	Learning quickly when irrelevant attributes abound: A new linear threshold algorithm - Littlestone - 1988
583	Queries and concept learning - Angluin - 1988
556	The weighted majority algorithm - Littlestone, Warmuth - 1994
347	Learning decision lists - Rivest - 1987
290	WebWatcher: A Learning Apprentice for the World Wide Web. AAAI Spring Symposium on Information Gathering from Heterogeneous, Distributed Environments - Armstrong, Freitag, et al. - 1995
267	How to use expert advice - Cesa-Bianchi, Freund, et al. - 1997
248	Efficient noise-tolerant learning from statistical queries - Kearns - 1998
211	Aggregating strategies - Vovk - 1990
169	Toward efficient agnostic learning - KEARNS, SCHAPIRE, et al. - 1994
155	On the Learnability of Boolean Formula - Kearns, Li, et al. - 1987
129	Universal prediction of individual sequences - Feder, Merhav, et al. - 1992
122	An analog of the minimax theorem for vector payoffs - Blackwell - 1956
122	Universal portfolios - Cover - 1991
117	Game theory, On-line Prediction and Boosting - Freund, Schapire - 1996
114	Empirical support for Winnow and weighted-majority based algorithms: results on a calendar scheduling domain - Blum - 1995
95	Redundant noisy attributes, attribute errors, and linear-threshold learning using Winnow - Littlestone - 1991
86	A game of prediction with expert advice - Vovk - 1998
71	Learning boolean functions in an infinite attribute space - Blum - 1992
65	Universal portfolios with side information - Cover, Ordentlich - 1996
64	Tracking the best disjunction - Auer, Warmuth - 1995
46	Learning in the presence of finitely or infinitely many irrelevant attributes - Blum, Hellerstein, et al. - 1995
44	Universal portfolios with and without transaction costs - Blum, Kalai - 1997
43	On-line prediction and conversion strategies - Cesa-Bianchi, Freund, et al. - 1994
40	On-line learning of linear functions - LITTLESTONE, LONG, et al. - 1995
35	Predicting a binary sequence almost as well as the optimal biased coin - Freund - 2003
31	On-line learning and the metrical task system problem - Blum, Burch - 1997
31	Learning Probabilistic Prediction Functions - DeSantis, Markowski, et al. - 1988
26	A randomization rule for selecting forecasts - Foster, Vohra - 1993
25	Asymptotically subminimax solutions of compound statistical decision problems - Robbins - 1951
21	Separating distribution-free and mistake-bound learning models over the boolean domain - Blum - 1994
14	Universal Sequential Learning and Decision from Individual Data Sequences - Merhav, Feder - 1992
13	On-line portfolio selection - Ordentlich, Cover - 1996
10	Toward e cient agnostic learning - Kearns, Schapire, et al. - 1992
8	Learning nested differences of intersection closed concept classes - Helmbotd, Sloan, et al. - 1989
7	Learning boolean functions in an in nite attribute space - Blum - 1992
5	Universal sequential coding of single measures - Shtarkov - 1987
4	Learning in the presence of nitely or in nitely many irrelevant attributes - Blum, Hellerstein, et al. - 1995
4	cient noise-tolerant learning from statistical queries - unknown authors - 1993
2	Learning nested di erences of intersection closed concept classes - Helmbold, Sloan, et al. - 1990
1	personal communication (a mistake-bound version of Rivest's decision-list algorithm - Littlestone - 1989