基于单参数的Lagrangian支持向量回归算法及其应用

A Single-Parameter Lagrangian Support Vector Regression and Its Application

支持向量机（Support Vector Machines，SVM）是基于统计学习理论框架下的一种处理非线性分类和非线性回归的有效方法。由于具有完备的理论基础和出色的学习性能，该方法已成为当前国际机器学习界的研究热点，能较好地对解决小样本、高维数、非线性和局部极小点等实际问题。提出了一种基于单参数的Lagrangian支持向量回归算法，并将该算法应用在外贸货物吞吐量预测中。估算结果证明了这种改进的支持向量回归算法在吞吐量预测中的有效性和实用性。

The Support Vector Machines （SVM） is an effective method of treating non-linearity classification and regression, based on the frame of statistical learning theory. Because of its complete theoretical background and excellent generalization performance, it has become the hotspot of machine learning in the world. This method can solve the practical problems such as limited samples, high dimensions, non-linearity and local minimum. A new algorithm of Support Vector Regression is proposed in this article, which is named Single-Parameter Lagrangian Support Vector Regression, and it is used in the prediction of the foreign-trade cargo throughputs of Shanghai Port. The results of experiments show the practicability and effectiveness of this modified algorithm in the prediction field of throughputs.