Learning binary relations and total orders

Author

Goldman, Sally A. ; Rivest, Ronald L. ; Schapire, Robert E.

Author_Institution

MIT Lab. for Comput. Sci., Cambridge, MA, USA

fYear

1989

fDate

30 Oct-1 Nov 1989

Firstpage

46

Lastpage

51

Abstract

The problem of designing polynomial prediction algorithms for learning binary relations is studied for an online model in which the instances are drawn by the learner, by a helpful teacher, by an adversary, or according to a probability distribution on the instance space. The relation is represented as an n×m binary matrix, and results are presented when the matrix is restricted to have at most k distinct row types, and when it is constrained by requiring that the predicate form a total order

Keywords

computational complexity; learning systems; matrix algebra; adversary; binary matrix; binary relations; distinct row types; helpful teacher; instance space; learner; learning; online model; polynomial prediction algorithms; predicate; probability distribution; total orders; Algorithm design and analysis; Animals; Bipartite graph; Computer science; Laboratories; Polynomials; Prediction algorithms; Predictive models; Probability distribution; Size measurement;

fLanguage

English

Publisher

ieee

Conference_Titel

Foundations of Computer Science, 1989., 30th Annual Symposium on

Conference_Location

Research Triangle Park, NC

Print_ISBN

0-8186-1982-1

Type

conf

DOI

10.1109/SFCS.1989.63454

Filename

63454