Knowledge discovery based on fuzzy rough set theory by inclusion degrees for space load forecasting

A spatial load forecasting (SLF) model realized by decision table is proposed, in which the setting of condition and decision attributes is based on fuzzy rough set (FRS) theory. The proposed SLF model has the capability of coping with not only natural geographic factors but also abrupt issues. Using data mining (DM) technology, a large amount of data related to land-use-change can be analyzed and managed in order to avoid the turbulence occurring in data management. In the model proposed, the problems investigated in this paper are as follows: 1. An iterated algorithm for progressive sampling based on ¿boundary region¿ of rough-set (RS) theory is developed; 2. Fuzzy division for problem region, by means of knowledge data in database (KDD) is proposed; 3. Attributes with continuous value is coped with to be convenient with information table based on FRS so that the output of SLF is better than that based on RS theory. The test results show that the proposed method is more feasible for SLF in complex situations than traditional methods.