作者: T. Dudok de Wit , E. Floriani
关键词:
摘要: A parametric method similar to autoregressive spectral estimators is proposed determine the probability density function (pdf) of a random set. The proceeds by maximizing likelihood pdf, yielding estimates that perform equally well in tails as bulk distribution. It therefore suited for analysis short sets drawn from smooth pdfs and stands out simplicity its computational scheme. Its advantages limitations are discussed.