Alternative combination of quantum immune algorithm and back propagation neural network

The handwritten recognition is an important but difficult problem and the back propagation neural network (BP neural network) is the common solution. However, considering the shortcoming that BP neural network can easily be trapped in the local optimal solutions and have slow convergence, this paper proposes a new method to improve the performance with the combination of BP neural network and the quantum immune algorithm (QIA). This method which we called AQICA-BP can accelerate the convergent rate and escape from the local optimal in time. Thus we apply the AQICA-BP to the handwritten recognition problems and the results prove its feasibility. Moreover, we compare it with both single BP and QIA-BP, it shows that the new model is desirable.