• Federated Learning(FL)
• Hierarchical Federated Learning(HFL)
• Convolutional neural network(CNN)

The thesis titled aims to assess the efficacy of hierarchical federated learning (HFL) setups in distributed machine learning scenarios, focusing on convolutional neural network (CNN) architectures. Hierarchical federated learning involves organizing devices into multiple tiers, each responsible for aggregating and refining model updates before passing them to higher tiers. The study evaluates various aspects such as communication overhead, model convergence rate, and overall accuracy achieved by different CNN models within the HFL framework. Through extensive experimentation and analysis, the research seeks to provide insights into the potential benefits and challenges of employing hierarchical structures in federated learning settings, contributing to the advancement of distributed machine learning methodologies.