Authors:

Aritro Chatterjee

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

Frustrated quantum systems-where competing spin interactions prevent all constraints from being satisfied-exhibit complex critical behavior that is difficult to classify using standard tools. Understanding phase transitions in such systems is crucial for condensed matter physics and quantum technologies, especially as materials like SrCu2(BO3)2 realize frustrated lattice models such as the Shastry- Sutherland lattice in the lab. This project explores whether supervised machine learning can classify quantum phases and universality classes from simulation data, even in the presence of disorder. We combine Density Matrix Renormalization Group (DMRG), a numerical method for finding ground states of quantum systems, with quantum annealing using D-Wave's hybrid solvers to study the Shastry-Sutherland and Triangular lattices. We extract correlation functions and magnetization profiles- including one-third magnetization plateaus in the Shastry Sutherland lattice-and use these to train machine learning models. Our simulations currently use Julia-based DMRG and simulated annealing; we expect to soon incorporate data from physical quantum annealers and begin full- scale ML training. By integrating machine learning with quantum simulation, this project aims to create scalable, automated tools to detect criticality and phase structure in frustrated systems, including the effects of defects. Such tools could accelerate experimental mapping of phase diagrams and inform the design of new quantum materials. Keywords: [no keywords provided]

Source:

Purdue University / 2025

Topics:

No topics listed

Co-authors:

Aritro Chatterjee