中文摘要：

在半参数模型补偿最小二乘估计中,如何有效地计算正则化参数是求解的关键。在建立一种新的补偿最小二乘模型的基础上,基于广义交叉核实与L曲线准则,采用最优化方法计算正则化参数。仿真计算表明,本文方法是一种计算正则化参数更为有效的方法。

英文摘要：

In penalized least squares estimation with the semi-parametric model,the key is how to calculate regularization parameter effectively.An optimization method with a new model of the penalized least squares,was proposed for calculating the regularization parameter based on criteria of the generalized cross validation and L curve.The results of simulation calculation show that themethod is more effective for calculating regularization parameter.