Robust test statistics induced from the minimum L2 distance estimation

被引:0
作者
Pak, RJ [1 ]
机构
[1] Taejon Univ, Dept Stat, Taejon 300716, South Korea
关键词
Hellinger distance; Hellinger deviance test; nonparametric density estimates; robustness;
D O I
10.1080/00949659908811976
中图分类号
TP39 [计算机的应用];
学科分类号
081203 ; 0835 ;
摘要
The test statistics are derived based on a minimum problem of L-2 distance between the model density and a corresponding density estimator when the null hypothesis is simple or composite involving subspace. The newly proposed tests are asymptotically efficient and robust against slight contamination on an underlying density. The proposed test statistics are easy enough to be handled without extensive knowledge of computing unlike some of the existing deviance tests.
引用
收藏
页码:193 / 207
页数:15
相关论文
共 17 条
  • [1] [Anonymous], 1979, The Advanced Theory of Statistics
  • [2] EFFICIENT ROBUST-TESTS IN PARAMETRIC MODELS
    BERAN, R
    [J]. ZEITSCHRIFT FUR WAHRSCHEINLICHKEITSTHEORIE UND VERWANDTE GEBIETE, 1981, 57 (01): : 73 - 86
  • [3] THE AUTOMATIC ROBUSTNESS OF MINIMUM DISTANCE FUNCTIONALS
    DONOHO, DL
    LIU, RC
    [J]. ANNALS OF STATISTICS, 1988, 16 (02) : 552 - 586
  • [4] DONOHO DL, 1983, FESTSCHRIFT EL LEHMA, P157
  • [5] FINSTERWALDER CE, 1976, J ASSOC OFF ANA CHEM, V59, P169
  • [6] Hampel F. R., 1986, ROBUST STAT APPROACH
  • [7] HAMPEL FR, 1977, T 7 PRAG C INF THEOR, VA, P187
  • [8] HINKELY DV, 1974, THEORETICAL STAT
  • [9] Neyman J, 1928, BIOMETRIKA, V20A, P175, DOI 10.2307/2331945
  • [10] RAO CR, 1948, P CAMB PHILOS SOC, V44, P50