Learning with queries corrupted by classification noise
Jeffrey C. Jackson, Eitan Shamir, Clara Shwartzman · 2002
Kearns introduced the "statistical query" (SQ) model as a general method for producing learning algorithms which are robust against classification noise. We extend this approach in several ways, in order to tackle algorithms that use "membership queries": focusing on the more stringent model of "persistent noise". The main ingredients in the general analysis are: (1) Smallness of dimension of both the targets' class and the queries' class. (2) Independence of the noise variables. Persistence restricts independence forcing repeated invocation of the same point x to give the same label. We apply the general analysis and ad-hoc considerations to get noise-robust version of Jackson's Harmonic Sieve (1995), which learns DNF under the uniform distribution. This corrects an error in his earlier analysis of noise tolerant DNF learning.