Analysis of Data Mining and Dynamic Neural Network for Data Prediction

Abstract

Data prediction, as an important symbol of network technology innovation and development, is a technical process of estimating future data by combining existing data. Nowadays, with the comprehensive development of mobile network and social network, people are faced with more and more electronic data in daily life. At this time, how to accurately predict future data or understand the development trend of data is of great significance to industry construction. Neural network, as a computational model built by computer to simulate the process of human brain neuron processing information, has certain nonlinear modeling ability in practical application, and can adapt to master the law of data hiding as soon as possible. Therefore, in this paper, the neural network model and fuzzy system are discussed in depth, and the fuzzy neural network model is chosen to analyze the data prediction, and a general prediction framework based on fuzzy C clustering and ANFIS hybrid learning algorithm is proposed in the practical research, and an improved fuzzy C clustering based on density weighting (IDWFCM) is proposed. The final simulation results show that the clustering effect of IDWFCM algorithm is not affected by noise data, so that the convergence speed of the system is higher than the traditional clustering algorithm, the overall increase of 60%, and the clustering accuracy also increases from 88.4% to 94.2%.