GA-BASED OPTIMIZED CNN NETWORK FOR IMAGE CLASSIFICATION

Abstract

In computer vision Convolutional Neural Networks (CNNs) have revolutionized the field of image classification and AI techniques have been used with them to solve complex problems. However, designing an optimal CNN architecture for a specific task can be a complex and time-consuming process due to the large number of hyper-parameters involved. This paper proposes a novel approach that utilizes Genetic Algorithms (GAs) with CNN to automatically optimize the architecture of a CNN network for image classification tasks to improve the performance in terms of accuracy. The standard Intel image classification dataset employed for assessment of execution to enhance the performance.