UNIVERSITY OF ST ANDREWS Statistics Seminars ________________________________________________________________________________ MONDAY 15 NOVEMBER at 4 p.m. in Lecture Theatre D of the Mathematical Institute Dr Nial FRIEL (University of Glasgow) "Bayesian Model selection for partially observed diffusion models" ABSTRACT:- In this talk we present a method to tackle this problem using data augmentation by treating the paths between observed points as missing data. For a fixed model formulation, the strong dependence between the missing paths and the volatility of the diffusion can be broken down by adopting the recently presented method of Roberts and Stramer (Biometrika, 2001). We describe how this method may be extended to the case of model selection via reversible jump MCMC. In addition we extend the formulation of a diffusion model to capture a potential non-Markov state dependence in the drift. Issues of appropriate choices of priors and efficient trans-dimensional proposal distributions for the reversible jump algorithm are also addressed. The approach is illustrated using simulated data and an example from finance. This work is in collaboration with Petros Dellaportas (Athens) and Gareth Roberts (Lancaster). MONDAY 29 NOVEMBER at 4 p.m in Lecture Theatre A of the Mathematical Institute Dr Sandra CATLIN (University of Nevada, Las Vegas - visiting University of Strathclyde). "Stochastic Modeling of Hematopoiesis" ABSTRACT:- This talk describes the application of stochastic modeling to understanding stem cell behavior. Stem cells reside in the bone marrow and are the cells from which all the constituents of blood derive. Hematopoiesis is the multi-stage process in which stem cells, through sequential division, differentiation, and maturation, give rise to all types of circulating blood cells. Because stem cells are difficult to identify, their behavior (e.g. rates of self-replication and differentiation) must be inferred from observations of partially differentiated progenitor cells. A two-compartment hidden Markov model has been proposed to describe samples of cells representative of this stage of development taken over time from female Safari cats. We review several methods of parameter estimation in the model, and the inherent difficulties involved. We show how the model can be used to gain insight into stem cell related diseases, such as chronic myelogenous leukemia, and possible therapeutic strategies. ________________________________________________________________________________ Tea will be available from 3.45 p.m. on both November 15 and November 29. Visitors will be very welcome. Further information from: Dr I B J Goudie email: [log in to unmask] ________________________________________________________________________________