ABSTRACT
This chapter provides a survey of techniques for inference about parameter values for multivariate exponential distributions. It focuses on the bivariate model for ease of exposition and indicates when multivariate extensions can be found. The chapter describes inference for the Marshall-Olkin model, and the bivariate model of Freund. It deals with the absolutely continuous bivariate exponential distribution of H. W. Block and A. P. Basu. The chapter looks at inference for a class of models induced by individuals in groups sharing unobserved common random effects. It discusses an estimation scheme based on the EM-algorithm. The chapter analyses results for one of E. J. Gumbel’s bivariate exponential model and for the bivariate model proposed by F. Downton.
