Study on Cost Forecast Method of Power Projects Based on Data Mining Technology

Abstract

Based on cost data of certain zone history power transmission line projects, applying data mining technology including relative analysis, clustering analysis and support vector machine theory, one kind of new cost forecast methods of power projects is presented. Firstly, applying relative analysis and partial correlation analysis in SPSS15.0 software package, cost forecast index system is built after simplifying technical conditions of power projects. Next, using, clustering analysis, noise information of cost data of history projects is deleted. At last, using support vector machine theory, cost forecast model of power projects is designed. Simulation results of real power transmission line projects in Matlab7.0 software package show such model is valid and feasible.

Material Consumption Estimation of High-Rise Buildings Based on SVM

“…SVM method is mostly used alone [10][11] , while a few may combine it with FIS (fuzzy inference system) [12] . Peng points out that SVM method has a significant advantage over other traditional methods in generalization ability and predict accuracy for small samples [13] .…”