Metropolis Hastings Calibration Excel Example

This spreadsheet is intended as an example to demonstrate how the Metropolis Hastings algorithm can be implemented within microsoft Excel to undertake Bayesian inference. If you are considering programming the Metropolis Hastings algorithm in another language/modelling package this example may be useful for you. This example uses a very simple state transition model (with 3 states) and uses data observations of persons in State B and persons moving to Stage C. This structure and approach can be extended to a larger more complex model and with more parameters and datasets. This spreadsheet may be a useful illustration of the process of the MH algorithm for those considering programming this algorithm in another package. Warning! This example is intended as a rough guide to the process only. For further details consult a statistics reference. Referencing The author has used a similar approach to calibrate a natural history model for colorectal cancer. The methods are published here: Whyte S, Walsh C, Chilcott J. Bayesian Calibration of a Natural History Model with Application to a Population Model for Colorectal Cancer. Medical Decision Making 2011;31:625-641. http://www.ncbi.nlm.nih.gov/pubmed/21127321