Inference for a two-component mixture of symmetric distributions under log-concavity
- Citation data:
Bernoulli, ISSN: 1350-7265, Vol: 24, Issue: 2, Page: 1053-1071
- Publication Year:
- Captures 3
- Readers 3
- Mendeley 3
In this article, we revisit the problem of estimating the unknown zero-symmetric distribution in a twocomponent location mixture model, considered in previous works, now under the assumption that the zerosymmetric distribution has a log-concave density. When consistent estimators for the shift locations and mixing probability are used, we show that the nonparametric log-concave Maximum Likelihood estimator (MLE) of both the mixed density and that of the unknown zero-symmetric component are consistent in the Hellinger distance. In case the estimators for the shift locations and mixing probability are √n-consistent, we establish that these MLE's converge to the truth at the rate nin the Ldistance. To estimate the shift locations and mixing probability, we use the estimators proposed by (Ann. Statist. 35 (2007) 224-251). The unknown zero-symmetric density is efficiently computed using the R package logcondens.mode.