Intelligent Data Transformation Based on Knowledge Based

Data transformation method is well known in Knowledge Discovery in Databases (KDD) process and data mining in order to transform raw data into concepts at higher levels concepts. A number of promising data transformation methods have been studied and developed. Despite the great advantages offered by these data transformation methods, these methods still requires further improvement. In order to handle data transformation problems, in this research, we propose an intelligent data transformation (IDT). It is developed by combining several intelligent techniques, i.e., rough set, neural networks, knowledge based and statistic. IDT method has been tested by using Jakarta Stock Exchange (JSX) data set. Our study shows that IDT method could (i) transform raw data into transformed data accurately and intelligently, and (ii) transform raw data into transformed data with different number of classes.