NEURO FUZZY APPROACH TO DATA CLUSTERING: A FRAMEWORK FOR ANALYSIS

Abstract

Clustering and pattern recognition have been used for various purposes since times immemorial. However, with the advancements in computing technology and information and communication technology clustering has achieved great significance in data analysis, data mining, knowledge management, artificial intelligence, control processes, image recognition, etc. The ocean of valuable data which is being added every day to the already mountains of data poses a big challenge to the world of computing, as it is practically impossible to handle such a large quantity of data individually. As this data contains valuable hidden information in it, therefore, needs to be understood not only to solve human problems but also to avoid many disasters for betterment of human welfare. Therefore, there is pressing need for better and faster methods of data management which starts with better methods of data storage. Data storage itself is a problem unless it is known where to store which data. Again cluster analysis comes to the rescue and allows data managers to store infinite data in finite or fewer clusters. While, on the one hand the clustering solves the problem of data storage, it also helps in extracting valuable information from the data through its analysis. The information in the data, however, can be deciphered through understanding sequences, associations, and patterns in the data, which needs better and faster learning methods for which fuzzy and neurofuzzy methods have emerged. Against this backdrop the current paper attempts to educate about this field and provides a framework for neurofuzzy cluster analysis.