Title: Estimation of a two-component semiparametric location-shifted mixture model
Authors: Jingjing Wu - University of Calgary (Canada) [presenting]
Weixin Yao - UC Riverside (United States)
Sijia Xiang - Zhejiang University of Finance and Economics (China)
Xiaofan Zhou - University of Calgary (Canada)
Abstract: Two efficient and robust estimators are discussed for a two-component semiparametric mixture model where the two components are unknown location-shifted symmetric distributions. Our estimators are derived by minimizing either the Hellinger distance (MHD) or the profile Hellinger distance (MPHD) between the model and a nonparametric density estimation. We propose simple and efficient algorithms to find the proposed MHD and MPHD estimators. Simulation studies are conducted to examine the finite sample performance of the proposed estimators and procedures and to compare them with other existing methods. We observe from our empirical studies that the two proposed estimators work very competitively with the existing methods for normal mixtures and much better for non-normal mixtures. More importantly, the proposed estimators are robust when data are contaminated with outlying observations. A real data is analyzed to illustrate the application of the proposed estimators.