Machine Learning the Many-Body Problem
-
Abstract: Condensed matter physics is the study of the collective behaviour of infinitely complex assemblies of interacting electrons, magnetic moments, atoms or qubits. This complexity is reminiscent of the “curse of dimensionality” commonly encountered in machine learning. Despite this curse, the machine learning community has developed techniques with remarkable abilities to classify, characterize and interpret complex sets of real-world data, such as images or natural languages. Here, we show that modern neural network architectures for supervised learning can be used to identify phases and phase transitions in a variety of condensed matter Hamiltonians. These neural networks can be trained to detect ordered states, as well as topological states with no conventional order, directly from raw state configurations sampled theoretically or experimentally. Further, such configurations can be used to train a stochastic variant of a neural network, called a Restricted Boltzmann Machine (RBM), for use in unsupervised learning applications. We show how RBMs can be sampled much like a physical Hamiltonian to produce configurations useful for estimating physical observables. Finally, we examine the power of RBMs for the efficient representation of classical and quantum Hamiltonians, and explore applications in quantum state tomography useful for near-term multi-quoit devices.
Date and Time
Location
Hosts
Registration
- Date: 29 Mar 2018
- Time: 02:00 PM to 03:00 PM
- All times are (GMT-08:00) America/Vancouver
-
Add Event to Calendar
- TRIUMF Auditorium
- 4004 Wesbrook Mall
- Vancouver, British Columbia
- Canada V6T 2A3
- Building: Main Office Building
- Contact Event Host
- Co-sponsored by TRIUMF
Speakers
Dr Roger Melko
Machine Learning the Many-Body Problem
Biography:
Dr. Melko is Canada Research Chair in Computational Many-Body Physics. Dr. Melko's research interests involve strongly-correlated many-body systems, with a focus on emergent phenomena, ground state phases, phase transitions, quantum criticality, and entanglement. He emphasizes computational methods as a theoretical technique, in particular the development of state-of-the-art algorithms for the study of strongly-interacting systems. Dr. Melko's work has employed Monte Carlo simulations and Density Matrix Renormalization Group methods to explore the low-temperature physics of classical and quantum magnetic materials, cold atoms in optical lattices, bosonic fluids and low-dimensional systems. He is particularly involved in studying microscopic models that display interesting quantum behavior in the bulk, such as superconducting, spin liquid, topological, superfluid or supersolid phases. He is also interested in broader ideas in computational physics, the development of efficient algorithms for simulating quantum mechanical systems on classical computers, and the relationship of these methods to the field of quantum information science.
Address:Department of Physics and astronomy, University Of Waterloo, Waterloo, Ontario, Canada, N2L 3G1