Research On Data Mining And Visualization Of Urban Ambient Air Quality

In the past years, computer techniques especially of database techniques have developed greatly, area of people's activities has been extended, and rhythm of life has speeded up. People are able to get and store data more quickly, easily and cheaply, which make the data and information increase exponentially. Facing the great capacity of data, people are under the pressure of "information explosion" and "data glut". It will be garbage if the massive data can't be exploited. Data mining which aims at extracting novel and useful knowledge from large volumes of data, has emerged rapidly in last decades, it integrates techniques of machine learning, statistic learning and database.Jinan urban environmental protection and monitoring station has accumulated a large amount of historical data of monitoring which has very important meanings to the analysis and prediction of whole urban ambient air quality. It is not easy for decision-makers to understand and use traditional data mining techniques and algorithms, people need new techniques to support the realization of the logical analysis and use on existing data. It is easy to understand data and results of data mining using visualization techniques; it can allow comparison and test of results. Visualization can be used to guide the data mining algorithms to verify the logical correctness of the business data sets.This paper has discussed the relevant theory and technique of ambient air quality data mining and visualization, then establishes an archetypal system based on it, and integrates the existing ambient air quality monitoring and management system of Jinan urban environmental protection and monitoring station.Firstly, the paper introduces the basic theory and related techniques of data mining and visualization systemically. Secondly, the paper gives a prediction model of the ambient air quality based on rough set and B-P neural network and to do the prediction of the ambient air quality. Thirdly, the paper introduces an implementation method of visualization of forecasting data using GIS proprietary methods and visualization techniques. Fourthly, the paper gives a method of visual data mining which is based on parallel coordinates and cluster analysis. Finally the paper provides a data mining and visualization archetypal system which is built in windows operating system, Visual C++ 6.0 and provides data preprocessing, B-P neural network for data mining, dynamic visualization of forecasting data and visual clustering of the urban ambient air quality.The archetypal system establishes the model and method of forecasting and evaluation of the urban ambient air quality; it can meet the request for urban ambient air quality analysis on the basis of massive data tentatively and be regarded as the decision supplementary means of administrative department of environmental protection. The research and development of this system has some innovative, it provides basic conditions and necessary means for the establishment of urban ambient air quality data management, analysis and evaluation systems with the international advanced level.