Leveraging AWS EMR for scalable and efficient neural network deployment in cloud computing

Abstract

This research investigates the potential of AWS Elastic MapReduce (EMR) as a cloud-based solution for building and deploying neural networks, focusing on its performance, scalability, and flexibility compared to traditional on-premises systems. The study offers a comparative analysis of stochastic neural networks deployed using AWS EMR Serverless, with financial data from Google Finance used to assess model performance through the Root Mean Square Error (RMSE) metric. The results highlight the advantages of cloudbased machine learning, particularly in handling large-scale datasets.