PCA-based PSO-BP neural network optimization algorithm

BP neural network inherits many disadvantages such as slow convergence speed and easily converging to local minimum. The input data generally has a high-dimensional feature. To improve the performance of neural network, we propose a novel algorithm. Before inputting the data into the neural network, this algorithm reduces the dimension of the data with PCA algorithm. Then, this algorithm simplifies the structure of the neural network and reduces the amount of computation combined with PSO-BP algorithm. Simulation results experiments demonstrate that the proposed algorithm improves the overall efficiency of neural networks, which proves that PCA-Based PSP-BP algorithm is better than PSO-BP algorithm.