Improved BP Neural Network Based On Ant Colony Algorithm And Its Application

Neural network models have been widely used in control and optimization,classification and prediction,pattern recognition and image processing.However,because the construction of the neural network model has a certain dependence on the choice of the initial connection weights and thresholds,the classification accuracy and stability of the neural network when it comes to big data classification problems are affected.This paper establishes a BP neural network optimization algorithm based on ant colony algorithm.It uses ant colony algorithm to optimize the performance in solving complex problems.It optimizes the weights and thresholds of neural network in the solution space iteratively to control the relative of the neural network model.Errors improve the stability and accuracy of the model.The improved neural network was applied to the classification of farmer households in rural mortgage loans,and compared with the traditional neural network classification results to provide a decision reference for rural financial institutions.、The main contents are as follows:(1)Establish neural network weights,threshold coding,path selection methods based on ant colony algorithm;(2)According to the characteristics of ant colony algorithm in the process of solving complex problems optimization and the basic principle of BP neural network algorithm,an improved BP neural network algorithm framework and structure with ant colony algorithm as the core of related parameter optimization is constructed.Determine related parameter settings and selections;(3)By classifying the rural financial data,comparing the traditional algorithm with the convergence speed of the algorithm,the accuracy of the classification,and the accuracy,the effectiveness and feasibility of the method in solving practical problems are examined.