Using neural networks and rough set to analyze Creativity

Williams Creativity Test B (WCTB) and Adolescent Scientific Creativity Scale (ASCS) were used to measure the creative affective and scientific creativity for 550 middle school students. In these students, 70% of them were used as modeling group, and the other as testing group. Generalized regression neural network (GRNN) and multivariable linear regression (MLR) were used for modeling and testing. The result showed the fitting error of GRNN model was lower than the error of MLR. In the decision table, creative affective factors (CAF) were used as conditional attributes, and scientific creativity (SC) was used as decision attribute. The data of CAF was discretized with SOM network. The attribute reduction was conducted with SDM. Eight rules were extracted. It was found that curiosity and risk-taking were very important for scientific creativity.