Learning in an Infinite Attribute Space 13 Kearns, M., Li, M., Pitt, L., and Valiant, L. (1987). Recent results on boolean concept learning. In Proceedings of the Fourth International Workshop on Machine Learning, pages 337--352, University of California, Irvine. Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
Avrim Blum - School Of Computer Self-citation (Space) (Correct)
....add at least 1 relevant attribute and at most n total attributes to S, so the largest number of attributes ever considered is nr, where r = jR f T j. Since there are at most r different Pm used, this procedure makes at most 2rM nr mistakes queries total. 6 The Halving Algorithm We now consider learning when computational constraints are ignored. Littlestone (1987) defines opt(C) to be the best possible worst case mistake bound achievable by any (not necessarily polynomial time) algorithm for learning class C. If Cm is a concept class over f0; 1g m , then we have opt(C m ) log 2 jC m j. This can be seen by using the standard Halving Algorithm which ....
Learning in an Infinite Attribute Space 13 Kearns, M., Li, M., Pitt, L., and Valiant, L. (1987). Recent results on boolean concept learning. In Proceedings of the Fourth International Workshop on Machine Learning, pages 337--352, University of California, Irvine.
Information Filtering: Selection Mechanisms In Learning Systems - Markovitch (1989) (25 citations) (Correct)
No context found.
Machine Learning, 3, 285-317. Kearns, M., Li, M., Pitt, L., & Valiant, L. G. (1987). Recent Results on Boolean Concept Learning. In Proceedings of Forth International Workshop on Machine Learning (pp. 337-352). Irvine, California: Morgan Kaufmann.
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