ON A KERNEL APPROACH TO A SCREENING PROBLEM

• 成果类型: Article
• 署名作者: BOYS, RJ
• 刊物名称: JOURNAL OF THE ROYAL STATISTICAL SOCIETY SERIES B-STATISTICAL METHODOLOGY
• ISSN/ISSBN: 1369-7412
• 发表日期: 1992
• 页码: 157-169
• 关键词: bivariate normal probabilities; calculating noncentral-t; integral useful; regression; prediction; variables; variate; models

摘要

Screening methods are developed which raise the proportion of individuals possessing an attribute in the retained population to a predetermined level. A specification region for the screening variable is derived by using a kernel smoothing approach. The sensitivity of the procedure is assessed within a range of parametric models. Comparisons are also made with empirical and estimative solutions. The kernel method is illustrated by an example of screening within a medical context.