ANFIS Data Mining Model Research And Application

The advent of the information explosion era indicates that big data will bring huge changes to our lives,so the work of data mining is particularly important.Data mining,literally,means digging from a large amount of data to regular and valuable information and knowledge that can be understood and applied.There are many types of data mining models,and neural network models are one of them.Species.Since the neural network technology was first proposed and developed until now,various types of neural networks have emerged and have been gradually applied to various fields,and the fuzzy neural network technology has provided new scientific problems for many scientific problems Solutions and ideas.The ANFIS model is an adaptive fuzzy neural network,which organically combines traditional neural networks and fuzzy logic to make up for the shortcomings of traditional neural networks.Since the 1990 s,scholars have begun to use neural networks to model meteorological forecasting and climate analysis.However,with the deepening of research,various problems have also begun to emerge.In When using a neural network model to forecast the meteorological,the initial weight of the neural network,the network structure,and the learning coefficients are difficult to determine,so a large amount of training is required to determine the network structure and the learning parameters of the network,but this will happen again In the case of overfitting,the generalization ability of the model will be affected.At present,the forecast time of weather forecast is short(up to 15 days),so this paper attempts to use ANFIS model to solve the problem of long-term weather forecast.The paper aims to predict some meteorological elements(temperature,precipitation,evaporation,etc.)for a long time.This paper has completed the following work:1)Based on domestic and foreign research,research the fuzzy control theory and combine it with traditional neural networks;2)Through in-depth research on the ANFIS neural network model,the ANFIS network model is realized;3)Explore the possibility of ANFIS neural network model in the field of meteorological forecast,design and implement a network system,and conduct experiments with the collected ground meteorological data.;4)Pre-process the meteorological data and extract features,and then build models from the three aspects of temperature,precipitation and evaporation, and evaluate the results,then demonstrate the feasibility of ANFIS in the field of long-term weather prediction.