Mathematical Sciences: Diagnostic Procedures for Robust Methods

回归诊断和稳健程序是统计学中的两个领域，在过去十年中显示出快速增长。我们将研究用稳健程序来判断模型拟合的诊断程序。由R和M估计、有界影响GR和GM估计以及高分解估计拟合的模型的残差的性质都将被研究。还将开发用于评估适合性的图形程序。

Regression diagnostics and robust procedures are two areas in statistics that have exhibited rapid growth in the last decade.Diagnostic procedures for judging fits of models fitted by robust procedures will be investigated. Properties of residuals from models fitted by R and M estimates, bounded influence GR and GM estimates, and high breakdown estimates, will all be studied. Graphical procedures for assessing the fit will be also developed.