ruth@utstat.uucp (Ruth Croxford) (01/16/90)
Topic: Bayesian linear regression with error terms that have symmetric unimodal densities Speaker: Lawrence J. Brunner Date: Thursday, Jan. 18, 1990 4-5:00 p.m. Place: Rm. 1085, Sidney Smith Hall, 100 St. George Street, U of T Abstract: Bayes methods are provided for a multiple linear regression model in which the error terms have densities that are symmetric and unimodal at zero, but whose form is otherwise unknown. The posterior distribution of the vector or regression coefficients is obtained. Simulations are presented in which the frequentist variability of the parameter estimates (posterior expectations) is investigated. When the error terms have a heavy-tailed density, the present method yields estimates of the true regression coefficients that are much more accurate than the least-squares estimates. ------------ Coffee and tea will be served in the De Lury Lounge (SS6006) at 3:30 p.m.