Image Classification Using Federated Averaging Algorithm

Machine learning privacy and data ownership con-cerns have prompted the exploration of innovative methods to address these issues. This research delves into the realm of image classification using federated averaging algorithms as a promising solution. By capitalizing on federated learning frameworks, multiple collaborating parties can collectively build a machine learning model without the need to share their raw data. This paper presents a comprehensive overview of the Federated Averaging algorithm, focusing on its application to image classification tasks. This approach can potentially be used in various applications where data privacy is a concern. The dataset had sixty thousand training photos and ten thousand test images, divided into ten categories. The categories are T-shirt, pant, dress, sweater, coat, sandal, shirt, sneakers, bag, and boots in each example, which was a 28×28 gray-scale image. The study showcases the effectiveness of Federated Averaging in image classification, revealing outcomes that rival those attained through traditional centralized training methods. With a notable achievement of 87.27% accuracy in image classification, this research highlights the practical viability of Federated Learning while upholding stringent data privacy standards.