More efficient PAC-learning of DNF with membership queries under the uniform distribution

作者:

Highlights:

摘要

An efficient algorithm exists for learning disjunctive normal form (DNF) expressions in the uniform-distribution PAC learning model with membership queries (J. Comput. System Sci. 55 (1997) 414), but in practice the algorithm can only be applied to small problems. We present several modifications to the algorithm that substantially improve its asymptotic efficiency. First, we show how to significantly improve the time and sample complexity of a key subprogram, resulting in similar improvements in the bounds on the overall DNF algorithm. We also apply known methods to convert the resulting algorithm to an attribute efficient algorithm. Furthermore, we develop a technique for lower bounding the sample size required for PAC learning with membership queries under a fixed distribution and apply this technique to produce a lower bound on the number of membership queries needed for the uniform-distribution DNF learning problem. Finally, we present a learning algorithm for DNF that is attribute efficient in its use of random bits.

论文关键词:Probably approximately learning,Membership queries,Disjunctive normal form,Uniform distribution,Fourier transform

论文评审过程:Received 3 December 2001, Revised 15 June 2003, Accepted 17 October 2003, Available online 19 December 2003.

论文官网地址:https://doi.org/10.1016/j.jcss.2003.10.002