Summary: On the Sample Complexity of Noise­Tolerant Learning
Javed A. Aslam #
Department of Computer Science
Dartmouth College
Hanover, NH 03755
Scott E. Decatur +
Laboratory for Computer Science
Massachusetts Institute of Technology
Cambridge, MA 02139
Abstract
In this paper, we further characterize the complexity of noise­tolerant learning in the PAC
model. Specifically, we show a general lower bound
of# # log(1/#)
#(1-2#) 2 # on the number of examples
required for PAC learning in the presence of classification noise. Combined with a result of
Simon, we e#ectively show that the sample complexity of PAC learning in the presence of
classification noise
is# # VC(F)
#(1-2#) 2 + log(1/#)
#(1-2#) 2 # . Furthermore, we demonstrate the optimality of the

Source: Aslam, Javed - College of Computer Science, Northeastern University

Collections: Computer Technologies and Information Sciences