Hybrid deep learning approaches on HPC and quantum computing for data analysis

Abstract

This thesis explores the transformative role of machine learning, especially deep learning (DL), in engineering simulations, using material science as a key application area. By transitioning from human-driven to computer-analyzed simulations, DL can accelerate simulation workflows and enhance data insights. However, the computational and storage demands of DL present challenges that quantum computing might address. This research investigates how hybrid workflows, combining DL with quantum neural networks (QNNs), can improve tasks such as image classification and partial differential equation (PDE) solving.