FUNDAMENTALS OF MAXENT IN DATA-ANALYSIS

Neutron diffraction studies, like many others, pose problems of inference. The proper tool for analysing these is probability calculus. We stress this point, and show how to incorporate a preference for "smooth" reconstructions into the probabilistic framework. This is illustrated with synthetic interpolation data. Smoothness is incorporated in a natural way, which is easily compatible with maximum entropy analysis. With a firm foundation in probability theory, and the inclusion of smoothness, we expect maximum entropy will become even more useful in the future.