固定尺度最小二乘支持向量机
Fixed Size Least Squares Support Vector Machines
-
摘要: 针对最小二乘支持向量机(LS-SVM)在进行回归预测时存在的稀疏性缺陷问题,采用固定尺度最小二乘支持向量机,即固定支持向量数量进行改进。仿真结果表明:固定尺度最小二乘支持向量机在训练各种样本数据集时,有效地避开了LS-SVM中的稀疏性问题,且训练速度快,同时具有良好的预测精度。Abstract: For the defect of sparseness in regression predicting with least squares support vector(machines),fixed size LS_-SVM is adopted,which evades the problem of sparseness in LS_-SVM and takes on fast training speed.The simulation results indicate that fixed size LS_-SVM shortens the training time enormously and possesses good predicting precision on different datasets.