Lower Bounds for Agnostic Learning via Approximate Rank |
| |
Authors: | Adam R Klivans Alexander A Sherstov |
| |
Affiliation: | 1. Department of Computer Sciences, The University of Texas at Austin, 1 University Station C0500, Austin, TX, 78712-0233, USA
|
| |
Abstract: | We prove that the concept class of disjunctions cannot be pointwise approximated by linear combinations of any small set of
arbitrary real-valued functions. That is, suppose that there exist functions f1, ?, fr\phi_{1}, \ldots , \phi_{r} : {− 1, 1}n →
\mathbbR{\mathbb{R}} with the property that every disjunction f on n variables has $\|f - \sum\nolimits_{i=1}^{r} \alpha_{i}\phi
_{i}\|_{\infty}\leq 1/3$\|f - \sum\nolimits_{i=1}^{r} \alpha_{i}\phi
_{i}\|_{\infty}\leq 1/3 for some reals a1, ?, ar\alpha_{1}, \ldots , \alpha_{r}. We prove that then $r \geq
exp \{\Omega(\sqrt{n})\}$r \geq
exp \{\Omega(\sqrt{n})\}, which is tight. We prove an incomparable lower bound for the concept class of decision lists. For the concept class of majority
functions, we obtain a lower bound of W(2n/n)\Omega(2^{n}/n) , which almost meets the trivial upper bound of 2n for any concept class. These lower bounds substantially strengthen and generalize the polynomial approximation lower bounds of Paturi
(1992) and show that the regression-based agnostic learning algorithm of Kalai et al. (2005) is optimal. |
| |
Keywords: | |
本文献已被 SpringerLink 等数据库收录! |
|