Bassist User's Guide for Version 0.83

Hannu Toivonen, Heikki Mannila, Jouni Seppänen, and Kari Vasko: Bassist User's Guide for Version 0.83. Report C-1999-36, Department of Computer Science and Rolf Nevanlinna Institute, University of Helsinki, June 1999. 21 pages. <http://www.cs.helsinki.fi/TR/C-1999/36>

Full paper: gzip'ed Postscript file
Metadata: XML file

Abstract

Bassist is a tool for Bayesian statistical analysis. Given a high-level specification of a full probability model, Bassist can be used to obtain an approximate sample from the joint posterior distribution of the model parameters and missing data. Bassist is based on Markov chain Monte Carlo (MCMC) techniques, especially the Metropolis-Hastings-Green method.

The model specification language of Bassist is declarative. The design of the language emphasizes clarity in the model description by separating between the model specification, the simulation details, and the particulars of the data. Within the Bassist system, the model specifications are first compiled into a C ++ program, which then reads the data and performs the actual simulation.

Bassist is publicly available on the World Wide Web at < http://www.cs.helsinki.fi/research/fdk/bassist/ >.

Index Terms

Categories and Subject Descriptors:
D.1.2 Automatic programming
G.3 Probability and Statistics
I.6.7 Simulation Support Systems
I.6.8 Types of Simulation: Monte Carlo

General Terms: Algorithms, Experimentation

Additional Key Words and Phrases: Bayesian probability models, MCMC simulation, statistical modeling