Explore machine learning applications in quantum simulation through this 46-minute lecture by Cornell University's Eun-Ah Kim at IPAM's Model Reduction in Quantum Mechanics Workshop. Delve into the historical context of quantum mechanics, from X-ray diffraction to projective measurements, before focusing on modern machine learning techniques for quantum emergence. Examine the architecture and training of neural networks, particularly Correlator Convolutional Neural Networks (CCNN), for interpreting image-like quantum data. Discover how unsupervised and supervised machine learning approaches are used to identify new phases in programmable Rydberg quantum simulator snapshots, including edge-ordering and rhombic phases. Gain insights into the application of supervised machine learning for understanding complex quantum correlations and its potential to revolutionize quantum simulation research.
Machine Learning for Quantum Simulation - IPAM at UCLA
Institute for Pure & Applied Mathematics (IPAM) via YouTube
Overview
Syllabus
Intro
TUNNELING DENSITY OF STATES, IN 1962
X-ray diffraction in 1913
Projective Measurements in 1922
Machine Learning Quantum Emergence
Train a neural network to recognize best hypothesis?
DETERMINED BY WEIGHTS AND BIASES
TRAINING THROUGH FEEDBACK
CORRELATOR CONVOLUTIONAL NEURAL NETWORKS: AN INTERPRETABLE ARCHITECTURE FOR IMAGE-LIKE
Convolutional Neural Networks (CNN)
The validity of CCNN's learning?
MACHINE LEARNING DISCOVERY OF NEW PHASES IN PROGRAMMABLE RYDBERG QUANTUM SIMULATOR SNAPSHOTS
SQUARE-LATTICE RYDBERG PHASES
Unsupervised Pass at the Phase Diagram
Supervised Learning
Supervised Phase Diagram
Entanglement in Striated Phase
New Phase l: Edge-ordering
New Phase II: Rhombic Phase
Machine Learning for Quantum Simulation Supervised ML: Learn characteristic Correlations NEW INSIGHT into COMPLEX DATA
Taught by
Institute for Pure & Applied Mathematics (IPAM)
