Quantum Boltzmann Machine Training via Imaginary-Time Evolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing quantum machine learning technologies face challenges in creating an effective training process for quantum Boltzmann machines that enables exact evaluation of accuracy, relying on approximations rather than precise methods.
Innovation Solution
The system employs quantum imaginary-time evolution and a Hadamard circuit to evaluate the Kullback-Leibler divergence gradient and Hessian through sampling procedures, allowing for precise training of quantum Boltzmann machines by generating samples and using these metrics to optimize parameter configurations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If quantum imaginary-time evolution and Hadamard circuit are used to evaluate Kullback-Leibler divergence gradient, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent introduces quantum imaginary-time evolution as an intermediary mechanism to transform the difficult-to-evaluate Kullback-Leibler divergence gradient into a form that can be measured through quantum sampling. The Hadamard circuit serves as another intermediary to enable the actual measurement of expectation values. These intermediaries bridge the gap between the theoretical gradient evaluation and practical quantum measurement, achieving precise gradient evaluation while managing circuit complexity.
2Productivity
If quantum sampling procedure is used for gradient evaluation, then productivity is improved, but difficulty of detecting and measuring increases
Solution Approach 1:
The patent replaces the classical mechanical sampling procedure with a quantum mechanical sampling approach. Instead of using classical random sampling to estimate gradients, the system uses quantum imaginary-time evolution to generate samples that directly reflect the underlying probability distribution. This substitution enables more efficient gradient evaluation through quantum parallelism, improving training productivity while the Hadamard circuit provides a systematic method to handle the measurement complexity.
Data Source
AI summary
Systems, computer-implemented methods, and computer program products to facilitate training of quantum Boltzmann machines by quantum imaginary-time evolution. According to an embodiment, a system can comprise computer executable components stored in memory. The computer executable components comprise an evaluation component that evaluates a Kullback-Leibler divergence gradient by a sampling procedure, where samples are generated by quantum imaginary-time evolution and a Hadamard circuit.