论文部分内容阅读
将组块3×2交叉验证方法用于高维回归中的调节参数选择.首先通过ISIS方法把模型的维数降低到样本个数以内,然后使用AENET方法对降维后的模型进行进一步的降维和参数估计,使用组块3×2交叉验证方法选择最佳的调节参数.综合考虑模拟实验中各种调节参数选择方法(AIC、BIC、EBIC、HBIC、5折交叉验证、组块3×2交叉验证)的EMSE值、方差以及计算复杂度,结果表明基于组块3×2交叉验证的方法是有其优势的.
The block 3 × 2 cross-validation method is applied to the selection of adjustment parameters in high-dimensional regression.At first, the dimension of the model is reduced to the number of samples by ISIS method, and then the AENET method is used to further reduce the dimensionality of the model The optimal adjustment parameters are selected by using 3 × 2 cross-validation method of chunks, and the method of selecting various adjustment parameters (AIC, BIC, EBIC, HBIC, 5-fold cross validation, Cross-validation) EMSE values, variance and computational complexity, the results show that the method based on chunk 3 × 2 cross-validation has its advantages.