Fuzzy Data Mining Technique In The Application Of The Atmospheric System

With the rapid development of information technology, data mining technology as an important subject has been more and more widely applied to different fields. However, for lots of fuzzy phenomena in life, data mining technology has a great limitation. Towards this problem, fuzzy data mining technology has been proposed forward.The most frequently used algorithm in data mining technology is Fuzzy Clustering Analysis. This paper introduced transitive closure method and fuzzy C-means algorithm, both of which are the most widely used methods in Fuzzy Clustering Analysis. But they both have advantages and disadvantages in solving real problems. Based on this, the paper introduced a mixed fuzzy clustering analysis algorithm, which is the combination of the above two algorithms. However, this algorithm still has some defects. Aiming at these defects, the paper has made some algorithm improvements:on the one hand, F statistics had been introduced to calculate the best classification number for clustering in order to make the cluster results more objective; on the other hand, when using C-mean method to develop softening score correction, we have made some innovation towards the computing algorithm of initiating Dividing Matrix and specified the algorithm which is fully realized by MATLAB programming. This improvement has also reduced the errors caused by human factors, making the results more reasonable and approachable to the practical situation.In this paper, based on making use of some air data in internship programs and combining the improved algorithm, we have divided the rainfall regions of the central China and compared it with other results with different algorithms. The facts have proved that the clustering result is very ideal and close to the actual situation.