Quantum neural network training method and device, electronic equipment and medium

Abstract

The invention provides a quantum neural network training method and device, electronic equipment, a computer readable storage medium and a computer program product, and relates to the field of computers, in particular to the technical field of quantum computers. According to the implementation scheme, L + 1 parameterized quantum circuits and L data coding circuits are determined; acquiring a plurality of training data pairs including independent variable data and dependent variable data; for each training data pair, alternately connecting parameterized quantum circuits and data coding circuits in series to form a quantum neural network, wherein the data coding circuits respectively code independent variable data in the training data pair; running the quantum neural network from an initial quantum state, and measuring the obtained quantum state to obtain a measurement result; calculating a loss function according to the measurement results corresponding to all the training data pairs and the corresponding dependent variable data; and adjusting to-be-trained parameters of the parameterized quantum circuit and the data coding circuit to minimize the loss function.

Description

technical field [0001] The present disclosure relates to the field of computers, in particular to the technical field of quantum computers, and in particular to a quantum neural network training method, device, electronic equipment, computer-readable storage medium, and computer program product. Background technique [0002] Many problems in daily production and life belong to the problem of function simulation, such as stock trend forecasting, weather forecasting, etc. With the development of artificial intelligence technology, Deep Neural Network (DNN) is widely used to solve the above problems. However, the DNN model requires a lot of parameters. For large-scale DNNs, the number of parameters is often hundreds of millions. In addition, the hyperparameters of the model are also difficult to adjust, and it is easy to overfit during training. [0003] With the rapid development of the field of quantum computing, recent quantum computing equipment can already support some sh...

AI Technical Summary

Problems solved by technology

However, the DNN model requires a lot of parameters. For large-scale DNNs, the number of parameters is often hundreds of millions. In addition, the hyperparameters of the model are also difficult to adjust, and it is easy to overfit during training.

Images