The objectives of this paper are to develop a class of semiparametric estimations and to find an appropriate estimator for this class. The idea is to multiply an initial parametric density estimate with a proposal correction function. We consider this study in the multidimensional case. A new estimator should be particularly useful when either parametric or nonparametric methods have problems. The comparisons with the traditional kernel estimator are presented by some practical examples. Mochamad Sonhaji; Metodologi Statistik, Badan Pusat Statistik (BPS)
