Showing 1–3 of 3 results for author: Jothishwaran, C A

On Learning with LAD

Authors: C. A. Jothishwaran, Biplav Srivastava, Jitin Singla, Sugata Gangopadhyay

Abstract: The logical analysis of data, LAD, is a technique that yields two-class classifiers based on Boolean functions having disjunctive normal form (DNF) representation. Although LAD algorithms employ optimization techniques, the resulting binary classifiers or binary rules do not lead to overfitting. We propose a theoretical justification for the absence of overfitting by estimating the Vapnik-Chervone… ▽ More The logical analysis of data, LAD, is a technique that yields two-class classifiers based on Boolean functions having disjunctive normal form (DNF) representation. Although LAD algorithms employ optimization techniques, the resulting binary classifiers or binary rules do not lead to overfitting. We propose a theoretical justification for the absence of overfitting by estimating the Vapnik-Chervonenkis dimension (VC dimension) for LAD models where hypothesis sets consist of DNFs with a small number of cubic monomials. We illustrate and confirm our observations empirically. △ Less

Submitted 28 September, 2023; originally announced September 2023.

A quantum algorithm to estimate the closeness to the Strict Avalanche criterion in Boolean functions

Authors: C. A. Jothishwaran, Abhishek Chakraborty, Vishvendra Singh Poonia, Pantelimon Stanica, Sugata Gangopadhyay

Abstract: We propose a quantum algorithm (in the form of a quantum oracle) that estimates the closeness of a given Boolean function to one that satisfies the ``strict avalanche criterion'' (SAC). This algorithm requires $n$ queries of the Boolean function oracle, where $n$ is the number of input variables, this is fewer than the queries required by the classical algorithm to perform the same task. We compar… ▽ More We propose a quantum algorithm (in the form of a quantum oracle) that estimates the closeness of a given Boolean function to one that satisfies the ``strict avalanche criterion'' (SAC). This algorithm requires $n$ queries of the Boolean function oracle, where $n$ is the number of input variables, this is fewer than the queries required by the classical algorithm to perform the same task. We compare our approach with other quantum algorithms that may be used for estimating the closeness to SAC and it is shown our algorithm verifies SAC with the fewest possible calls to quantum oracle and requires the fewest samples for a given confidence bound. △ Less

Submitted 25 November, 2022; originally announced November 2022.

A quantum algorithm to estimate the Gowers $U_2$ norm and linearity testing of Boolean functions

Authors: C. A. Jothishwaran, Anton Tkachenko, Sugata Gangopadhyay, Constanza Riera, Pantelimon Stanica

Abstract: We propose a quantum algorithm to estimate the Gowers $U_2$ norm of a Boolean function, and extend it into a second algorithm to distinguish between linear Boolean functions and Boolean functions that are $ε$-far from the set of linear Boolean functions, which seems to perform better than the classical BLR algorithm. Finally, we outline an algorithm to estimate Gowers $U_3$ norms of Boolean functi… ▽ More We propose a quantum algorithm to estimate the Gowers $U_2$ norm of a Boolean function, and extend it into a second algorithm to distinguish between linear Boolean functions and Boolean functions that are $ε$-far from the set of linear Boolean functions, which seems to perform better than the classical BLR algorithm. Finally, we outline an algorithm to estimate Gowers $U_3$ norms of Boolean functions. △ Less

Submitted 30 June, 2020; originally announced June 2020.