Learning of Neural Networks with Quantum Computers and Learning of Quantum States with Graphical Models
Explore a cutting-edge presentation on quantum computing and machine learning applications. Delve into the Quantum Alphatron algorithm, which offers a polynomial speedup for kernelized regression and two-layer neural network learning while maintaining theoretical guarantees. Examine the learning of quantum states using graphical models, focusing on states approximated by neural network quantum states represented by restricted Boltzmann machines (RBMs). Discover robustness results for efficient provable two-hop neighborhood learning algorithms applied to ferromagnetic and locally consistent RBMs, enabling certain quantum states to be learned with exponentially improved sample complexity compared to naive tomography. Gain insights into the intersection of quantum computing, machine learning, and computational learning theory in this 52-minute talk presented at IPAM's Mathematical Aspects of Quantum Learning Workshop.