초록 열기/닫기 버튼
In this study, we propose the multi-block L-type estimator in symmetric location model. We also derive the asymptotic properties of the proposed estimator by well-known approximating the best weight function. We calculate the asymptotic relative efficiency (ARE) of the proposed estimator when the number of block is two. The ARE of the proposed estimator is calculated compared to Rao-Cramer lower bound in case of normal and Cauchy distribution, and also compared to Pitman value in double exponential case. In addition, we conduct the Monte Carlo study to find out the performance of the proposed estimator. From the Monte Carlo study our proposed estimator works pretty good in all distributions and especially works better in light-tailed and heavy-tailed distributions we studied. In robust location problem, M-type estimators like Huber and Tukey type and the trimmed mean with reasonable trimming portion are the most popular methods. From the Monte Carlo study, we can't say our proposed estimator is better than well known M-estimators, but we think it can be used in alternative purposes.
