An improved BP neural network algorithm and its application

In order to overcome the disadvantages such as such as low convergence speed, falling into local minimum easily, bad generalization ability of BP neural network algorithm, the paper presents a new wavelet neural network and applies it to evaluate university PE teaching. First, the improved algorithm uses wavelet algorithm to redesign and simplify the algorithm structure of BP algorithm; Second, genetic algorithm and BFGS algorithm are used to speed up the convergence of BP algorithm and calculation flows are redesigned; Finally, based on constructing evaluation indicator system of PE teaching, the improved algorithm is applied to evaluate PE teaching and experimental results indicate that the improved algorithm can improve evaluation accuracy and algorithm efficiency, decrease calculation time and can be used for evaluating other complicated systems practically.