This Paper investigates population I star samples from the viewpoint of the velocity distributions, each being viewed as a mixture. To obtain information on the class centers, clustering techniques are first applied. Secondly, we use a parametric maximum likelihood formulation that we solve by Redner-Walker's E.M. algorithm. The obtained results are compared and discussed from an astrophysical viewpoint.
Comparing Parametric and Nonparametric Statistical Methods for Studying the Velocity Distributions of Population I Stars
