A Novel Process Neural Networks Model Based on Quantum Computing

This work is a research on integrating quantum computing with process neural networks. To enhance the approximation and generalization ability of process neural networks (PNN), by studying the quantum implementation of information processing of process neuron, a new designing idea of process neuron, based on the quantum rotation gates and the multi-qubits controlled-Hadamard gates, is proposed in this paper. In the proposed approach, the discrete inputs are represented by the qubits, which, as the control qubits of the controlled-Hadamard gates after being rotated by the quantum rotation gates, control the target qubits for reverse. The model outputs are described by the probability amplitude of state  1 | in the target qubits. Then the quantum-inspired process neural networks (QPNN) are designed by applying the quantuminspired process neurons to the hidden layer and the classical neurons to the output layer. The algorithm of QPNN is derived by employing the principles of quantum computing and the {\it Levenberg-Marquardt} algorithm. Simulation results of a benchmark problem show that, under a certain condition, the QPNN is obviously superior to the classical PNN. Keywords-quantum computation, quantum rotation gates, multi-qubits controller-hadamard gates, quantum-inspired process neuron, quantum-inspired process neural networks