In this paper, we consider the issue of variable selection in partial linear single-index models under the assumption that the vector of regression coefficients is sparse. We apply penalized spline to estimate the non...In this paper, we consider the issue of variable selection in partial linear single-index models under the assumption that the vector of regression coefficients is sparse. We apply penalized spline to estimate the nonparametric function and SCAD penalty to achieve sparse estimates of regression parameters in both the linear and single-index parts of the model. Under some mild conditions, it is shown that the penalized estimators have oracle property, in the sense that it is asymptotically normal with the same mean and covariance that they would have if zero coefficients are known in advance. Our model owns a least square representation, therefore standard least square programming algorithms can be implemented without extra programming efforts. In the meantime, parametric estimation, variable selection and nonparametric estimation can be realized in one step, which incredibly increases computational stability. The finite sample performance of the penalized estimators is evaluated through Monte Carlo studies and illustrated with a real data set.展开更多
Disparities between the in situ and satellite values at the positions where in situ values are obtained have been the main handicap to the smooth modeling of the distribution of ocean chlorophyll. The blending techniq...Disparities between the in situ and satellite values at the positions where in situ values are obtained have been the main handicap to the smooth modeling of the distribution of ocean chlorophyll. The blending technique and the thin plate regression spline have so far been the main methods used in an attempt to calibrate ocean chlorophyll at positions where the in situ field could not provide value. In this paper, a combination of the two techniques has been used in order to provide improved and reliable estimates from the satellite field. The thin plate regression spline is applied to the blending technique by imposing a penalty on the differences between the satellite and in situ fields at positions where they both have observations. The objective of maximizing the use of the satellite field for prediction was outstanding in a validation study where the penalized blending method showed a remarkable improvement in its estimation potentials. It is hoped that most analysis on primary productivity and management in the ocean environment will be greatly affected by this result, since chlorophyll is one of the most important components in the formation of the ocean life cycle.展开更多
基金Supported by the National Natural Science Foundation of China(No.11671096)
文摘In this paper, we consider the issue of variable selection in partial linear single-index models under the assumption that the vector of regression coefficients is sparse. We apply penalized spline to estimate the nonparametric function and SCAD penalty to achieve sparse estimates of regression parameters in both the linear and single-index parts of the model. Under some mild conditions, it is shown that the penalized estimators have oracle property, in the sense that it is asymptotically normal with the same mean and covariance that they would have if zero coefficients are known in advance. Our model owns a least square representation, therefore standard least square programming algorithms can be implemented without extra programming efforts. In the meantime, parametric estimation, variable selection and nonparametric estimation can be realized in one step, which incredibly increases computational stability. The finite sample performance of the penalized estimators is evaluated through Monte Carlo studies and illustrated with a real data set.
文摘Disparities between the in situ and satellite values at the positions where in situ values are obtained have been the main handicap to the smooth modeling of the distribution of ocean chlorophyll. The blending technique and the thin plate regression spline have so far been the main methods used in an attempt to calibrate ocean chlorophyll at positions where the in situ field could not provide value. In this paper, a combination of the two techniques has been used in order to provide improved and reliable estimates from the satellite field. The thin plate regression spline is applied to the blending technique by imposing a penalty on the differences between the satellite and in situ fields at positions where they both have observations. The objective of maximizing the use of the satellite field for prediction was outstanding in a validation study where the penalized blending method showed a remarkable improvement in its estimation potentials. It is hoped that most analysis on primary productivity and management in the ocean environment will be greatly affected by this result, since chlorophyll is one of the most important components in the formation of the ocean life cycle.
