Research And Application Of Quantum Neural Network

Quantum Neural Network(QNN)is a research field that combines traditional Artificial Neural Network(ANN)with quantum knowledge.It is of great significance to enhance the potentialities of ANN.In this paper,we propose three QNN models based on the ANN model and combine them with quantum theory and quantum mechanics concepts such as quantum parallel computing,quantum gate circuits and variable quantum circuits,and the typical application of QNN is the field of network attack detection.The research is summarized in the following three aspects.A QGA-QGCNN model for quantum gate circuit neural network(QGCNN)optimization using quantum genetic algorithm(QGA)is proposed.QGCNN is constructed by quantum rotation gate and multi-qubit controlled-NOT gate,and the characteristic of fast global convergence of QGA to initialize QGCNN model parameters.From the simulation results,QGA-QGCNN is superior to ANN and ANN optimized by GA in the convergence performance and detection accuracy.In this paper,a variational quantum neural network(VQNN)model is put forward:a quantum-classical hybrid scheme consisting of quantum circuits combined with machine learning strategies.The quantum circuits comprised two parts:quantum state encoding circuits and variational quantum circuits(VQC)which simulates the classical probability distribution of the learning target and encodes the information into the quantum circuits parameters.Eventually the classical probability output distribution is acquired by measuring the VQC quantum state output and post-processed using a classical computer.This structure allows VQC to be quite easily combined with classical ML.Furthermore,this paper improves the original VQNN by introducing an intermediate measurement operation to enhance the nonlinearity of the model and designing two VQCs for different processing objectives.In addition,this paper explores the use of the VQNN as well as the improved VQNN to build classifiers based on practical applications and validated improved VQNN on a quantum processor.The test results suggest that the optimized model has high performance,and is higher than other classic contrast detection models,QGA-QGCNN and the original VQNN model.In addition,the model can be deployed in most recent noisy intermediate-scale quantum devices.