ZHAN tangsen, LIU weijie

(School of Information Engineering, Jingdezhen Ceramic Institute, Jingdezhen 33403)

Abstract: A combined predication algorithm was produced based on optimal AHP of high dimensional data, by which the partition of influencing factors and weight values were obtained through correlation coefficient calculation. Better predication results were gained with the determination of parameter weight value vector and optimal nonlinear problems. The prediction algorithm can overcome the subjective definition in paired comparison matrix and consistency test, and is capable of wide application.

Keywords: partition, vector of parameter weight value, optimal nonlinear problem