作者: R. K. Bryan
DOI: 10.1007/978-94-009-0683-9_12
关键词: Algorithm 、 Mathematics 、 Entropy (information theory) 、 Data point 、 Posterior probability 、 Cholesky decomposition 、 Variables 、 Principle of maximum entropy 、 Mathematical optimization 、 Scattering 、 Singular value decomposition
摘要: A numerical algorithm for the solution of Classic Maximum Entropy problem is presented, use when data are considerably oversampled, so that amount independent information they contain very much less than actual number points. Examples problems which this particularly appropriate dynamic light scattering, scattering and fibre diffraction. The application a general purpose entropy maximisation program then comparatively inefficient. In new variables in singular space transform between map (or image or spectrum) data, fewer either reconstruction. This reduction dimension allows direct evaluation posterior probability solution, thus enables ‘Classic Maxent’ to be solved completely.