Maximum Entropy and Bayesian Methods

1. Front Matter

   Pages i-xi

   PDF

2. Applications

   1. Flow and Diffusion Images from Bayesian Spectral Analysis of Motion-Encoded NMR Data

      • E. J. Fordham, D. Xing, J. A. Derbyshire, S. J. Gibbs, T. A. Carpenter, L. D. Hall

      Pages 1-12

   2. Bayesian Estimation of MR Images from Incomplete Raw Data

      • G. J. Marseille, R. de Beer, M. Fuderer, A. F. Mehlkopf, D. van Ormondt

      Pages 13-22

   3. Quantified Maximum Entropy and Biological EPR Spectra

      • S. M. Glidewell, B. A. Goodman, J. Skilling

      Pages 23-30

   4. The Vital Importance of Prior Information for the Decomposition of Ion Scattering Spectroscopy Data

      • R. Fischer, W. von der Linden, V. Dose

      Pages 31-40

   5. Bayesian Consideration of the Tomography Problem

      • W. von der Linden, K. Ertl, V. Dose

      Pages 41-49

   6. Using MaxEnt to Determine Nuclear Level Densities

      • N. J. Davidson, B. J. Cole, H. G. Miller

      Pages 51-58

   7. A Fresh Look at Model Selection in Inverse Scaterring

      • Vincent A. Macaulay, Brian Buck

      Pages 59-67

   8. The Maximum-Entropy Method in Small-Angle Scattering

      • Steen Hansen, Jürgen J. Müller

      Pages 69-78

   9. Maximum Entropy Multi-Resolution EM Tomography by Adaptive Subdivision

      • Li-He Zou, Zhengrong Wang, Louis E. Roemer

      Pages 79-89

   10. High Resolution Image Construction from IRAS Survey — Parallelization and Artifact Suppression

       • Yu Cao, Thomas A. Prince

       Pages 91-99

   11. Maximum Entropy Performance Analysis Of Spread-Spectrum Multiple-Access Communications

       • F. Solms, P. G. W. van Rooyen, J. S. Kunicki

       Pages 101-108

   12. Noise Analysis in Optical Fibre Sensing: A Study using the Maximum Entropy Method

       • L. Stergioulas, A. Vourdas, G. R. Jones

       Pages 109-116

3. Algorithms

   1. Autoclass — A Bayesian Approach to Classification

      • John Stutz, Peter Cheeseman

      Pages 117-126

   2. Evolution Review Of BayesCalc, A Mathematica™ Package for doing Bayesian Calculations

      • Paul Desmedt, Ignace Lemahieu, K. Thielemans

      Pages 127-134

   3. Bayesian Inference for Basis Function Selection in Nonlinear System Identification using Genetic Algorithms

      • Visakan Kadirkamanathan

      Pages 135-142

   4. The meaning of the word “Probability”

      • Myron Tribus

      Pages 143-155

   5. The Hard Truth

      • Kenneth M. Hanson, Gregory S. Cunningham

      Pages 157-164

   6. Are the Samples Doped — If so, How Much?

      • Anthony J. M. Garrett

      Pages 165-174

   7. Confidence Intervals from one Observation

      • C. C. Rodríguez

      Pages 175-182

This volume records papers given at the fourteenth international maximum entropy conference, held at St John's College Cambridge, England. It seems hard to believe that just thirteen years have passed since the first in the series, held at the University of Wyoming in 1981, and six years have passed since the meeting last took place here in Cambridge. So much has happened. There are two major themes at these meetings, inference and physics. The inference work uses the confluence of Bayesian and maximum entropy ideas to develop and explore a wide range of scientific applications, mostly concerning data analysis in one form or another. The physics work uses maximum entropy ideas to explore the thermodynamic world of macroscopic phenomena. Of the two, physics has the deeper historical roots, and much of the inspiration behind the inference work derives from physics. Yet it is no accident that most of the papers at these meetings are on the inference side. To develop new physics, one must use one's brains alone. To develop inference, computers are used as well, so that the stunning advances in computational power render the field open to rapid advance. Indeed, we have seen a revolution. In the larger world of statistics beyond the maximum entropy movement as such, there is now an explosion of work in Bayesian methods, as the inherent superiority of a defensible and consistent logical structure becomes increasingly apparent in practice.

• Markov model
• Probability theory
• classification
• image processing
• maximum entropy method
• neural networks
• thermodynamics

• Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Cambridge, England

  John Skilling, Sibusiso Sibisi

Policies and ethics