首页> 外文会议>Symposium on the interface >Bayesian Penalized Splines in Semi-Parametric Modeling
【24h】

Bayesian Penalized Splines in Semi-Parametric Modeling

机译:贝叶斯惩罚半导体造型中的曲目

获取原文

摘要

Penalized regression splines provide a useful tool for fitting complicated models with smooth components. Because they are sieve estimators, parametric tools such as likelihood and information criteria can be used for fitting. In this paper, I will demonstrate applications of Bayesian Penalized Splines to self-modeling regression and to varying coefficient models. We will also see that some statistics that are pivotal for the fixed knots (parametric) case do not appear to be pivotal for the sieve estimator. Also, knot placement, which has been shown to be of minimal importance for univariate smoothing, can have a large effect in more complicated settings.
机译:惩罚回归样条提供了一种有用的工具,用于使用平滑组件拟合复杂模型。因为它们是筛子估计,所以可以使用诸如可能性和信息标准的参数工具来配合。在本文中,我将展示贝叶斯惩罚样条的应用与自建模回归和变化系数模型。我们还将看到针对固定结的一些统计数据(参数)案例似乎似乎不适合筛分估计器。此外,已显示为单变量平滑的显着重要性的结放置可以在更复杂的环境中具有很大的效果。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号