An Integrated Intelligent Fuzzy System for Data Mining

This paper presents an advanced model based on fuzzy inference system, namely Standard Additive Model (SAM) for forecasting the output of any record given the input variables only from the database, the age of abalone in particular. The modeling and learning power of the SAM have been beneficial for the building of a model that is capable of prediction functionalities. In order to intensify the processing speed and accuracy of the SAM as well as to allow it to be more concise, we propose a learning process incorporate a genetic algorithm. In addition, advanced algorithms for unsupervised and supervised learning are employed. Experimental results have demonstrated that the system has the ability necessary for application to financial data mining and prediction in comparison with multilayer perceptron neural networks with the same context. In particular, the proposed method has shown a robustness against noise when the genetic algorithm is employed with a reasonable generation number.