ECCC-Report TR08-091https://eccc.weizmann.ac.il/report/2008/091Comments and Revisions published for TR08-091en-usTue, 21 Feb 2012 02:43:33 +0200
Revision 1
| On The Power of Membership Queries in Agnostic Learning |
Vitaly Feldman
https://eccc.weizmann.ac.il/report/2008/091#revision1We study the properties of the agnostic learning framework of Haussler (1992)and Kearns, Schapire and Sellie (1992). In particular, we address the question: is there any situation in which membership queries are useful in agnostic learning?
Our results show that the answer is negative for distribution-independent agnostic learning and positive for agnostic learning with respect to a specific marginal distribution. Namely, we give a simple proof that any concept class learnable agnostically by a distribution-independent algorithm with access to membership queries is also learnable agnostically without membership queries. This resolves an open problem posed by Kearns et al. (1992). For agnostic learning with respect to the uniform distribution over $\{0,1\}^n$ we show a concept class that is learnable with membership queries but computationally hard to learn from random examples alone (assuming that one-way functions exist).
Tue, 21 Feb 2012 02:43:33 +0200https://eccc.weizmann.ac.il/report/2008/091#revision1
Paper TR08-091
| On The Power of Membership Queries in Agnostic Learning |
Vitaly Feldman
https://eccc.weizmann.ac.il/report/2008/091We study the properties of the agnostic learning framework of Haussler (1992)and Kearns, Schapire and Sellie (1992). In particular, we address the question: is there any situation in which membership queries are useful in agnostic learning?
Our results show that the answer is negative for distribution-independent agnostic learning and positive for agnostic learning with respect to a specific marginal distribution. Namely, we give a simple proof that any concept class learnable agnostically by a distribution-independent algorithm with access to membership queries is also learnable agnostically without membership queries. This resolves an open problem posed by Kearns et al. (1992). For agnostic learning with respect to the uniform distribution over $\{0,1\}^n$ we show a concept class that is learnable with membership queries but computationally hard to learn from random examples alone (assuming that one-way functions exist).
Mon, 06 Oct 2008 14:47:39 +0200https://eccc.weizmann.ac.il/report/2008/091