| Michael Frazier and C. David Page. Learnability of recursive, nondeterminate theories: Some basic results and techniques. In Proceedings of the Third International Workshop on Inductive Logic Programming, Bled, Slovenia, 1993. |
....examples is not always less than m ff for any ff 1. 34 solved in any way that yields a concise hypothesis. A number of further questions suggest themselves. The learnability of recursive logic programs is a challenging problem; some results in this area appear in [Cohen, 1993c; Cohen, 1994b; Frazier and Page, 1993] The learnability of multiple clause predicate definitions is largely an open issue; although analysis is difficult, continued progress on the learnability of fairly general classes of DNF is encouraging [Flammini et al. 1992; Hancock, 1993; Kushilevitz and Roth, 1993] The learnability of ....
Michael Frazier and C. David Page. Learnability of recursive, nondeterminate theories: Some basic results and techniques. In Proceedings of the Third International Workshop on Inductive Logic Programming, Bled, Slovenia, 1993.
....1: Summary of learnability results for recursive programs standard equivalence query model) or a single k ary recursive clause plus a base case (if a basecase oracle is allowed) All other programs recursive programs are either cryptographically hard, or as hard as boolean DNF. 6 Related work Frazier and Page [1993a; 1993b] perform a similar analysis of the learnability of recursive programs with function symbols but without background knowledge. The only positive results they provide are for program classes that satisfy the following property: given a set of positive examples S that requires that all clauses ....
Michael Frazier and C. David Page. Learnability of recursive, non-determinate theories: Some basic results and techniques. In Proceedings of the Third International Workshop on Inductive Logic Programming, Bled, Slovenia, 1993. 52
....The second is a non recursive form, that is, consisting of two atoms. The following concept has this form. Technically, several other forms are possible. But these are less interesting, and their treatment here would only obscure the fundamental ideas. Please see [ Frazier and Page, 1993a; Frazier and Page, 1993b; Frazier, 1994 ] for additional details. 21 Example 8 Nonrecursive Concept p(fgf(X) p(hff(Y ) This concept classi es a ground atom as positive just if the ground atom is an instance of one of the two atoms in the concept. The third form is that of a contracting concept, as follows. ....
....] It is not known whether this class is pac predictable. But it is known that if the predicate symbol arities in this class are restricted to be at most k, for any xed k, and if variables are stationary, then the class is learnable by extended equivalence queries [ Frazier and Page, 1993a; Frazier and Page, 1993b ] This restricted language is denoted by H 2;k . We now brie y outline how the learning algorithm for H 2;1 can be used to predict H 2;k for any xed k. Let the following be a concept in H 2;k . p( 1 (X 1 ) k (X k ) p( 1 (Y 1 ) k (Y k ) p( 1 (Y 1 ) k (Y k ) ....
[Article contains additional citation context not shown here]
M. Frazier and C. D. Page. Learnability of recursive, nondeterminate theories: Some basic results and techniques. In Proceedings of the Third International Workshop on Inductive Logic Programming, pages 103-126, Ljubljana, Slovenia, 1993. J. Stefan Institute Technical Report IJS-DP-6707.
....The second is a non recursive form, that is, consisting of two atoms. The following concept has this form. 4 Technically, several other forms are possible. But these are less interesting, and their treatment here would only obscure the fundamental ideas. Please see [ Frazier and Page, 1993a, Frazier and Page, 1993b, Frazier, 1994 ] for additional details. 20 Example 8 Nonrecursive Concept p(fgf(X) p(hff(Y ) This concept classifies a ground atom as positive just if the ground atom is an instance of one of the two atoms in the concept. The third form is that of a contracting concept, as follows. ....
....1993b] It is not known whether this class is pac predictable. But it is known that if the predicate symbol arities in this class are restricted to be at most k, for any fixed k, and if variables are stationary, then the class is learnable by extended equivalence queries [Frazier and Page, 1993a, Frazier and Page, 1993b] This restricted language is denoted by H 2;k . We now briefly outline how the learning algorithm for H 2;1 can be used to predict H 2;k for any fixed k. Let the following be a concept in H 2;k . p(ff 1 (X 1 ) ff k (X k ) p(fl 1 (Y 1 ) fl k (Y k ) p(fi 1 (Y 1 ) fi k (Y k ....
[Article contains additional citation context not shown here]
M. Frazier and C. D. Page. Learnability of recursive, nondeterminate theories: Some basic results and techniques. In Proceedings of the Third International Workshop on Inductive Logic Programming, pages 103--126, Ljubljana, Slovenia, 1993. J. Stefan Institute Technical Report IJS-DP-6707.
No context found.
Frazier, M., & Page, C. D. (1993b). Learnability of recursive, non-determinate theories: Some basic results and techniques. In Proceedings of the Third International Workshop on Inductive Logic Programming Bled, Slovenia.
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