Image Classification Method using Dynamic Quantum Inspired Genetic Algorithm

TL;DR

This paper introduces a dynamic quantum-inspired genetic algorithm that uses quantum principles to improve feature selection for image classification, achieving near-perfect accuracy and outperforming traditional genetic algorithms.

Contribution

The paper proposes a novel D-QIGA with adaptive mechanisms and chromosome strategies, enhancing exploration and avoiding local optima in feature selection.

Findings

Achieves over 99.99% classification accuracy

Outperforms traditional Genetic Algorithms

Effective on benchmark and real-world problems

Abstract

This study presents a dynamic Quantum-Inspired Genetic Algorithm (D-QIGA) for feature selection, leveraging quantum principles like superposition and rotation gates to enhance exploration and exploitation. D-QIGA introduces adaptive mechanisms and a lengthening chromosome strategy to avoid local optima and improve optimization. Tested on benchmark and real-world problems, it significantly outperforms traditional Genetic Algorithms, achieving over 99.99% classification accuracy compared to GA's 95%.