:::

News & Events

:::

Speaker:Prof. Chun-Shu Chen (Institute of Statistics, NCU)

Seminar
Poster:Webmasters事件日期:2019-11-08
 
Topic:Local variable selection criterion based on a prediction perspective
 
Speaker:Prof. Chun-Shu Chen (Institute of Statistics, NCU)
 
Date Time:FRI. Nov 8,2019, 10:40 AM - 11:30 AM 
 
Place: 4F-427, Assembly Building I
 
Abstract
 
Variable selection and spatial prediction both are important issues in spatial statistics. If spatially varying means exist among different subareas, globally fitting a spatial regression model for the study area may be not suitable. To alleviate deviations from model assumptions, we propose a local variable selection criterion to locally select variables for each subarea. The proposed local criterion considers the global spatial dependence of observations and the characteristics of each subarea are also identified. It results in a composite spatial predictor which not only provides a more accurate spatial prediction, but also reduces the prediction variance. Statistical inferences of the proposed methodology are justified both theoretically and numerically.
 
Last modification time:2019-10-24 AM 10:13

  • recruiting animation-EN
cron web_use_log