Estimation and Classification for Finite Mixture Models Under Ranked-Set Sampling

• Activity Number: 640
• Type: Topic Contributed
• Date/Time: Thursday, August 8, 2013 : 8:30 AM to 10:20 AM
• Sponsor: Section on Nonparametric Statistics
• Author(s): Armin Hatefi*+ and Mohammad Jafari Jozani and Djemel Ziou
• Companies: University of Manitoba and University of Manitoba and Universit ´e de Sherbrooke
• Keywords: Ranked Set Samples ; Order Statistics ; Finite Mixture Models ; Expectation Maximization Algorithm ; Classification ; Latent Variables

Abstract

We consider maximum likelihood estimation of parameters of a finite mixture model (FMM) for independent order statistics data arising from ranked set sampling, as well as classification of the observed data. Proposing a ranked-based sampling design from a FMM, we develop new expectation-maximization algorithm resulting in more efficient estimators and more precise classifications of the observations than their counterparts based on a simple random sampling technique. Theoretical results are augmented with simulation studies.