Machine Learning Enabled Large-scale Modeling of Functional Electron Materials

The emergence of complex, functional electron materials has generated significant interest due to their potential applications in various technologies, including quantum computing and energy storage. However, understanding and modeling these materials present a major challenge because of the complexity of their electronic structure and interactions. This dissertation explores the use of machine learning to enable large-scale modeling of functional electron materials, with a focus on the Falicov-Kimball model, itinerant electron magnets, and disordered spin systems. We leverage neural network architectures to create efficient energy models, thereby enabling simulations that are otherwise computationally prohibitive. Our results demonstrate that ML can capture complex phenomena, such as phase separation and spin dynamics, with high accuracy. This work not only provides insights into the behavior of functional electron materials but also establishes a framework for employing ML in computational physics to solve otherwise intractable problems.