Lower bounds for linear decision lists

Chattopadhyay, Arkadev ; Mahajan, Meena ; Mande, Nikhil ; Saurabh, Nitin (2020) Lower bounds for linear decision lists Chicago Journal of Theoretical Computer Science, 26 (1). pp. 1-15. ISSN 1073-0486

[img] PDF
283kB

Official URL: http://doi.org/10.4086/cjtcs.2020.001

Related URL: http://dx.doi.org/10.4086/cjtcs.2020.001

Abstract

We demonstrate a lower bound technique for linear decision lists, which are decision lists where the queries are arbitrary linear threshold functions. We use this technique to prove an explicit lower bound by showing that any linear decision list computing thefunction MAJ ◦ XOR requires size 20.18n. This completely answers an open question of Turán and Vatan [19]. We also show that the spectral classes PL1,PL∞, and the polynomialthreshold function classes PTc1,PT1, are incomparable to linear decision lists.

Item Type:Article
Source:Copyright of this article belongs to Department of Computer Science, The University of Chicago.
Keywords:Linear threshold functions; Decision lists; Threshold circuits
ID Code:128014
Deposited On:14 Oct 2022 11:27
Last Modified:09 Nov 2022 09:40

Repository Staff Only: item control page