Research On Quantum Neural Network Model And Its Application To ECG Classification

Since Quantum Neural Network was proposed, it has achieved significant research results. Quantum neural network is the product of the combination of quantum computation and artificial neural network, it takes full advantage of the parallel feature of quantum computing, quantum superposition and quantum entanglement properties. Thereby, compared with the inherent defects of traditional artificial neural network, such as slow convergence speed and being easy to fall into local minimum, Quantum neural network has a faster convergence rate and higher classification ability than classical neural network. Therefore, It is greatly significant that quantum neural network is applied to ECG classification.Firstly, quantum neuron model and its features was studied. Neurons are the basic computing unit of the neural network, based on the study of classical neuron and traditional quantum neuron, a new quantum neuron was proposed and the characteristics of the new quantum neuron was investigated with comparison to the other quantum neurons.Secondly, the quantum neural network model and relative characteristics were studied. The new quantum neuron which was proposed in paper was used to construct a three layers feed-forward quantum neural network model. The new Quantum Neural Network adopted hyperbolic tangent function as activation function, By the way of function approximation, Quantum Neuron Network has a stronger approximation ability than the past quantum neural network and classical BP neural network.Finally, the new quantum neural network model was applied to the ECG classification. Using wavelet transform techniques to extract ECG feature data, and a method of ECG classification based on the new quantum neural network was proposed. The classification result of the extracted ECG feature parameters shows that Quantum Neuron Network has a more accurate classification result than traditional methods.