Chow参数问题的近最优解与半空间的低权逼近
Nearly optimal solutions for the Chow Parameters Problem and low-weight approximation of halfspaces
摘要:
布尔函数$f: \{-1,1\}^n \to \{-1,1\}$的\emph{Chow parameters}是其$n+1$ 0度和1度傅立叶系数。自1961年(Chow,Tannenbaum)以来,已知任何线性阈值函数$f$的Chow参数(的精确值)在所有布尔函数的空间内唯一地指定$f$,但直到最近(O 'Donnell和Servedio),还不知道\emph{reconstructing} $f$的有效算法(精确或近似)来自其Chow参数的精确或近似值。我们把这个重建问题的\emph{Chow Parameters Problem.}我们的主要结果是一个新的算法的周参数问题,给定(足够准确的近似)的周参数的任何线性阈值函数$f$,运行在时间$\tilde{O}(n^2)\cdot (1/\eps)^{O(\log^2(1/\eps))}$,并以高概率输出一个表示的LTF $f'$是$\eps$ -接近$f$。唯一以前的算法(O 'Donnell和Servedio)有运行时间$\poly(n) \cdot 2^{2^{\tilde{O}(1/\eps^2)}}.$作为我们的方法的副产品,我们表明,对于任何线性阈值函数$f$在$\{-1,1\}^n$,有一个线性阈值函数$f'$是$\eps$ -接近$f$,并具有所有的权重是整数最多$\sqrt{n} \cdot (1/\eps)^{O(\log^2(1/\eps))}$。这显著地改进了Diakonikolas和Servedio给出$\poly(n) \cdot 2^{\tilde{O}(1/\eps^{2/3})}$权重界的最佳先前结果,并且接近于$\max\{\sqrt{n},$ $(1/\eps)^{Ω(\log \log (1/\eps))}\}$(Goldberg,Servedio)的已知下界。我们的技术也产生了学习理论中相关问题的改进算法。
英文摘要:
The \emph{Chow parameters} of a Boolean function $f: \{-1,1\}^n \to \{-1,1\}$ are its $n+1$ degree-0 and degree-1 Fourier coefficients. It has been known since 1961 (Chow, Tannenbaum) that the (exact values of the) Chow parameters of any linear threshold function $f$ uniquely specify $f$ within the space of all Boolean functions, but until recently (O'Donnell and Servedio) nothing was known about efficient algorithms for \emph{reconstructing} $f$ (exactly or approximately) from exact or approximate values of its Chow parameters. We refer to this reconstruction problem as the \emph{Chow Parameters Problem.} Our main result is a new algorithm for the Chow Parameters Problem which, given (sufficiently accurate approximations to) the Chow parameters of any linear threshold function $f$, runs in time $\tilde{O}(n^2)\cdot (1/\eps)^{O(\log^2(1/\eps))}$ and with high probability outputs a representation of an LTF $f'$ that is $\eps$-close to $f$. The only previous algorithm (O'Donnell and Servedio) had running time $\poly(n) \cdot 2^{2^{\tilde{O}(1/\eps^2)}}.$ As a byproduct of our approach, we show that for any linear threshold function $f$ over $\{-1,1\}^n$, there is a linear threshold function $f'$ which is $\eps$-close to $f$ and has all weights that are integers at most $\sqrt{n} \cdot (1/\eps)^{O(\log^2(1/\eps))}$. This significantly improves the best previous result of Diakonikolas and Servedio which gave a $\poly(n) \cdot 2^{\tilde{O}(1/\eps^{2/3})}$ weight bound, and is close to the known lower bound of $\max\{\sqrt{n},$ $(1/\eps)^{Ω(\log \log (1/\eps))}\}$ (Goldberg, Servedio). Our techniques also yield improved algorithms for related problems in learning theory.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。