Research On Applications Of Data Mining Based On Rough Set Theory

Data mining is a new technology for drawing knowledge from data in databases. As an effective approach to processing incomplete, imprecise or uncertain information, rough set, with a powerful capability of qualitative analysis, has been playing an important role in the area of data mining. Data mining based rough set is one of important new parts of intelligence system.In order to deal with multidimensional, incomplete and understandable data, data mining based on rough set and its related concepts was introduced in detail. Firstly, the rise and development of rough set and its application in the field of data mining was chiefly studied. Secondly, the process of rough set was given. This algorithm reduces data from data warehouse. It enhances the capability of processing data. Finally, in order to deal with continuous attributes, a variety of discrete method was discussed.Because of the presence of large incomplete, noise, vague and random data, a rough set based neural network was presented. This algorithm reduces data from data warehouse by using rough sets. Thus, the number of the input variables and neurons is gotten. A kind of neural networks using data mining based on rough set with good understandability, simple computation and exact accuracy can be constructed.The approach was used to build a soft sensor model for estimating the temperature of the tank in batch stirred tank reactor (BSTR) and good results were obtained. It is shown that the approach can reduce the training time, improve the learning efficiency and enhance the predication accuracy.Data mining based rough set is a good data processing method. It can provide useful information for modeling, control, and other research work. Because it can improve the learning efficiency, enhance the predication accuracy, it is very helpful to real-time applications.