| Fuzzy clustering is a clustering analyzes method, which can be widely used in data analysis, knowledge discovery, pattern recognition, image processing and intelligent decision-making, and so on. Fuzzy clustering is usually employed to solve the problem that the things’property boundaries are not clear, while the classical clustering algorithms can’t handle it. Fuzzy clustering has been widely studied in academic and industrial areas, and favorable results and substantial benefits have been achieved.It is worth nothing that the most existing fuzzy clustering methods process the simple samples (employ a vector to describe with s features) while the related works on processing complex samples (the elements of the vector contain certain functions) are lacked. Moreover, with the rapid development of IT technology, the growth of data scale and diversity of data structure are far beyond the scope of traditional fuzzy clustering analysis. How to effectively complete the fuzzy clustering analysis and process complex samples becomes an urgent problem. Fortunately, the big data processing technology provides the basis to solve the above problems.The main works of this paper are as follows:(1) Study of fuzzy clustering under the Hadoop. First of all, the paper studies the algorithms of fuzzy clustering parallelized by Mapreduce, and explores the details of implementing the processes Map, Combiner, Reduce. Compared with other methods, the parallelized FCM algorithm is better.(2) Research of applications by fuzzy clustering based on big data processing. A new method is provided. First, the complex samples are pre-treated by the results of big data processing to determine the functions in the complex samples; then the FCM based on Mapreduce is used to complete data analysis, moreover, optimization objective rules are given. The experiments demonstrate that this method can efficiently and accurately complete data analysis.In addition, this method provides a solution for a class of similar problems. |
