According to the limitation of single kernel, support vector machine based on mixture kernels was proposed to improve the generalization ability and the precision of models.

• 摘要针对核函数方法中单个核函数的局限性，提出混合核支持向量机建模方法以提高模型的泛化能力和精度。