中国科技论文
中國科技論文
중국과기논문
Sciencepaper Online
2014年
7期
779-783
,共5页
特征向量%稀疏%支持向量%弓网系统
特徵嚮量%稀疏%支持嚮量%弓網繫統
특정향량%희소%지지향량%궁망계통
feature vector%sparse%support vector%pantograph-catenary system
提出基于特征向量选择(feature vector selection,FVS)的稀疏最小二乘支持向量机(sparse least squares support vector machine,SLS-SVM)模型,解决最小二乘支持向量机(least squares support vector machine,LS-SVM)稀疏化问题。采用FVS在特征空间构建特征向量子集,对训练样本进行稀疏线性重构;将稀疏化的特征向量作为支持向量,从而实现对LS-SVM稀疏化建模。将SLS-SVM模型进行弓网系统的仿真对比实验,结果表明SLS-SVM模型在取得高预报精度的同时,可实现支持向量的高度稀疏化,从而加快模型预报速度。
提齣基于特徵嚮量選擇(feature vector selection,FVS)的稀疏最小二乘支持嚮量機(sparse least squares support vector machine,SLS-SVM)模型,解決最小二乘支持嚮量機(least squares support vector machine,LS-SVM)稀疏化問題。採用FVS在特徵空間構建特徵嚮量子集,對訓練樣本進行稀疏線性重構;將稀疏化的特徵嚮量作為支持嚮量,從而實現對LS-SVM稀疏化建模。將SLS-SVM模型進行弓網繫統的倣真對比實驗,結果錶明SLS-SVM模型在取得高預報精度的同時,可實現支持嚮量的高度稀疏化,從而加快模型預報速度。
제출기우특정향량선택(feature vector selection,FVS)적희소최소이승지지향량궤(sparse least squares support vector machine,SLS-SVM)모형,해결최소이승지지향량궤(least squares support vector machine,LS-SVM)희소화문제。채용FVS재특정공간구건특정향양자집,대훈련양본진행희소선성중구;장희소화적특정향량작위지지향량,종이실현대LS-SVM희소화건모。장SLS-SVM모형진행궁망계통적방진대비실험,결과표명SLS-SVM모형재취득고예보정도적동시,가실현지지향량적고도희소화,종이가쾌모형예보속도。
A new model of sparse least squares support vector machine (SLS-SVM)is proposed to solve the sparseness problem of least squares support vector machine (LS-SVM),based on feature vector selection (FVS)method.A subset of feature vectors is defined in feature space to reconstruct all the training samples linearly.The sparse feature vectors are used as support vectors to model LS-SVM.SLS-SVM is simulated with pantograph-catenary system.It is shown that SLS-SVM can improve forecast preci-sion,and accelerate prediction speed by achieving highly sparse support vectors.