In general the model type is judged empirically from the available data. Since quantile models are designed to adapt to extreme values we would expect it to do better for heavy-tails distribution. The widespread and successful use of the Wakeby model is due to certain general characteristics of the distribution. A number of well-recognized methods can be used for estimating the unknown parameters. We have found that TheMLmethod has given results which are decidedly better than those obtained by other methods.
Si affronta il problema della stima di verosimiglianza di una variabile casuale espressa attraverso la sua funzione quantile. In particolare si segure l’approccio sviluppato per i dati raggruppati in classi. L’efficacia del metodo è illustrata su due casi concreti. Si adopera poi delle simulazioni per confrontarne l’efficienza con quella dei momenti pesati
Fitting Wakeby model using maximum likelihood
TARSITANO, Agostino
2005-01-01
Abstract
In general the model type is judged empirically from the available data. Since quantile models are designed to adapt to extreme values we would expect it to do better for heavy-tails distribution. The widespread and successful use of the Wakeby model is due to certain general characteristics of the distribution. A number of well-recognized methods can be used for estimating the unknown parameters. We have found that TheMLmethod has given results which are decidedly better than those obtained by other methods.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.