Robust and efficient estimation by minimising a density power divergence

• 成果类型: Article
• 署名作者: Basu, A; Harris, IR; Hjort, NL; Jones, MC
• 署名单位: Indian Statistical Institute; Indian Statistical Institute Kolkata; Northern Arizona University; University of Oslo; Open University - UK
• 刊物名称: BIOMETRIKA
• ISSN/ISSBN: 0006-3444
• DOI: 10.1093/biomet/85.3.549
• 发表日期: 1998
• 页码: 549559
• 关键词: minimum hellinger distance; PARAMETRIC MODELS; tests

摘要

A minimum divergence estimation method is developed for robust parameter estimation. The proposed approach uses new density-based divergences which, unlike existing methods of this type such as minimum Hellinger distance estimation, avoid the use of nonparametric density estimation and associated complications such as bandwidth selection. The pro; posed class of 'density power divergences' is indexed by a single parameter alpha which controls the trade-off between robustness and efficiency. The methodology affords a robust extension of maximum likelihood estimation for which alpha = 0. Choices of alpha near zero afford considerable robustness while retaining efficiency close to that of maximum likelihood.