基于核主元分析和支持向量机的f-CaO含量预测

Predicting the Free Calcium Oxide Content on the Basis of KPCA and Support Vector Machines

将核主元分析和支持向量机相结合,运用核主元分析对数据样本进行非线性特征提取,得到更易于回归的特征主元分量,达到了降低支持向量机的输入空间维数,然后运用最小二乘支持向量机进行训练,通过网格搜索和交叉验证确定最小二乘支持向量机的最优参数。建立了预测水泥熟料游离氧化钙含量的核主元分析支持向量机模型。计算结果表明提出的模型能有效地预测水泥熟料游离氧化钙含量。

Kernel principal component analysis（KPCA） and support vector machines（SVM） were combined in this study,which employed KPCA to conduct nonlinear feature extraction from the data sample and obtained feature principal components that are easier for regression operations.The number of input space dimensions that could lower the SVM has been met.After that,training was conducted by using the least squares support vector machines（LS-SVM） and determined the optimal parameters of the LS-SVM by means of grid searching and cross validation.A KPCA-SVM-based model was then established to predict the free calcium content in the clinker.Finally,our calculation results proved that the model proposed in this study can effectively predict the free calcium content in the clinker.