Extraction of Embedded Image Segment Data Using Data Mining with Reduced Neurofuzzy Systems

Abstract

To realize or implement the large dimensional image data, it may be taking a longer search time to detect the desired target. Recently, for the large amount of data and information in engineering or biomedical applications, various techniques including soft-computing techniques such as neural networks, fuzzy logic, or genetic algorithms, and multivariate analysis techniques like factor analysis, principal component analysis, or clustering analysis, are developed to extract the reduced meaningful information or knowledge from the original raw data. In this paper, for mining or diminishing the large dimension of the given raw image data, factor analysis, principal component analysis, and clustering analysis are used to make a model using fuzzy logic or neurofuzzy systems, which are applied to predict the characteristics of the images with reduced dimensions. Generally the procedure can produce more precise and reasonable results with reduced dimensions in order to predict the desired images. In addition, all those techniques are useful for searching and saving time for the desired images. Thus, the proposed techniques intend to propose hybrid systems with integrating various multivariate analysis techniques together to establish neurofuzzy or fuzzy logic systems to construct a reasoning system with more accurate and efficient.