本文中，主要是比較模糊類別最大概似法(fuzzy classification maximum likelihood，簡稱FCML)和EM演算法對伽瑪迴歸模型參數估計的準確性及效率。透過數值模擬的結果，FCML演算法僅在某些特定條件下有較好的結果，整體而言，兩種演算法各有優劣。

In this paper, we propose a approach called a fuzzy class model for Gamma regression, in the analysis of count data. On the basis of fuzzy classification maximum likelihood (FCML) procedures we be used an FCML algorithm for fuzzy class Gamma regression models. Traditionally, the EM algorithm had been used for latent class regression models. Thus, the accuracy and effectiveness of EM and FCML algorithms for estimating the parameters are compared. The results show that the proposed FCML algorithm and EM algorithm have advantages and disadvantages to regression analysis for count data.

[1] J.C.Bezdek (1981) Pattern recognition with fuzzy objective function algorithms. Plenum , New York

[2] A.P.Dempster, N.M.Laird, D.B.Rubin (1977) Maximum likelihood from
incomplete data via the EM algorithm (with discussion). J R Stat Soc B 39:1-38

[3] M.S.Yang (1993) A survey of fuzzy clustering. Math Comput Model 18: 1-16

[4] M.S.Yang (1993) On a class of fuzzy classification maximum likelihood
procedures. Fuzzy Sets Syst 57: 365-375

[5] M.S.Yang, C.Y.Lai (2005) Mixture Poisson regression models for
heterogeneous count data based on latent and fuzzy class analysis. Soft Comput 9:
519-524

[6] L.A.Zadeh (1965) Fuzzy sets. Information Control 8: 338-353