Abstract

This study aims too classify Guppy fish species using a Convolutional Neural Network (CNN) approach based on the pretrained VGG16 model. The dataset consists of Guppy fish images divided into training and validation sets. Data augmentation techniques, including rotation, shifting, zooming, and horizontal flipping, were applied to enhance the model’s generalization capabilities. The base layers of VGG16 were frozen during training to retain the pretrained weights from ImagNet. Evaluation results demonstrated high accuracy on both training dan validation data, supported by a Cofusion Matrix that illustrates the classification performance across various classes.