• 主题
高级搜索  >
历史搜索 清空历史
首页> 美国政府科技报告 >Probability Based Prediction from Measurements Using Gaussian Interpolation.Includes Appendices
【24h】

Probability Based Prediction from Measurements Using Gaussian Interpolation.Includes Appendices

机译：基于概率的高斯插值测量预测。包括附录

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

This report is concerned with doing predictions on data collected frommeasurements. In come cases the Gaussian Interpolation method may be an useful approach. It is a nonlinear interpolation method in a multidimensional space. This space contains supporting or data points. These supporting points need not be space regular. The estimation is based on the values of the nearby supporting points. These values are weighted according to the probability that the estimated pointed is equal to each of the supporting points. This method is in fact a table interpolation method which provides the most probable function value. Smoothing factors filter out the possible noise on the supporting points. By its nature this method always provides a stable estimation for any point: wild guesses are not possible. Gaussian interpolation will provide a smoothed and average prediction on the data. This induces that not all data points are predicted exactly. When this is desired it is possible calculate correction values to adjust the prediction. These correction values also provide some extrapolation capability.

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