Research on Dynamic Mining of Meteorological Data Based on PSO Algorithm

Thereis a fuzzy correlation between massive meteorological data, and this fuzzy correlation is difficult to determine. Therefore, a dynamic mining method for massive meteorological data based on fuzzy association rules is studied. This article combines EMD and MIC to design a time series data denoising algorithm for denoising massive meteorological data. Build a GAN-TRTI missing value completion function based on generative adversarial networks and time indices to fill in missing values in time series. Design a massive data dynamic mining algorithm using fuzzy association rules and particle swarm optimization algorithm to achieve dynamic mining of massive meteorological data. The test results show that in the mining results of the designed method, the average confidence of the samples in the mining rules is relatively high, ultimately stabilizing at around 92%, and the average support finally reaches 90%, indicating that the mining effect of the method is good.