作者: Panchi Li , Hong Xiao
DOI: 10.1007/S10489-013-0447-3
关键词:
摘要: To enhance the approximation and generalization ability of classical artificial neural network (ANN) by employing principles quantum computation, a quantum-inspired neuron based on controlled-rotation gate is proposed. In proposed model, discrete sequence input represented qubits, which, as control qubits after being rotated rotation gates, target qubit for rotation. The model output described probability amplitude state |1? in qubit. Then with (QNNSI) designed neurons to hidden layer layer. An algorithm QNNSI derived Levenberg---Marquardt algorithm. Experimental results some benchmark problems show that, under certain condition, obviously superior ANN.