An Efficient Algorithm for Maximum Tsallis Entropy using Fenchel-duality
نویسندگان
چکیده
We derive a dual-primal recursive algorithm based on the Fenchel duality framework, extending Dykstra’s successive projections and Csiszar’s I-projections schemes, to handle Tsallis MaxEnt. The Tsallis entropy Sq(p) is a one-parameter extension of Shannon’s entropyH(p) in the sense that Sq→1(p) = H(p). The solution of the Tsallis MaxEnt falls under a q-deformed Gibbs distribution which is a power-law distribution which contains the Gibbs distribution as a special case. The algorithm allows one to fit an extended distribution model to observations, which goes beyond the exponential distribution, yet is as simple as the original algorithms it extends.
منابع مشابه
Duality for maximum entropy diffusion MRI
We derive an entropy model for diffusion MRI and show that Fenchel duality techniques make this model tractable. This generalizes a model proposed by Daniel Alexander in 2005, in which the displacement of particles is confined to a sphere. In order to better suit the physics of the diffusion process, we propose to relax this constraint. The Kullback-Leibler relative entropy is used to measure t...
متن کاملCopulas with maximum entropy
We shall find a multi-dimensional checkerboard copula of maximum entropy that matches an observed set of grade correlation coefficients. This problem is formulated as the maximization of a concave function on a convex polytope. Under mild constraint qualifications we show that a unique solution exists in the core of the feasible region. The theory of Fenchel duality is used to reformulate the p...
متن کاملTsallis Maximum Entropy Lorenz Curves
In this paper, at first we derive a family of maximum Tsallis entropy distributions under optional side conditions on the mean income and the Gini index. Furthermore, corresponding with these distributions a family of Lorenz curves compatible with the optional side conditions is generated. Meanwhile, we show that our results reduce to Shannon entropy as $beta$ tends to one. Finally, by using ac...
متن کاملDuality of Maximum Entropy and Minimum Divergence
We discuss a special class of generalized divergence measures by the use of generator functions. Any divergence measure in the class is separated into the difference between cross and diagonal entropy. The diagonal entropy measure in the class associates with a model of maximum entropy distributions; the divergence measure leads to statistical estimation via minimization, for arbitrarily giving...
متن کاملPredicting Shear Stress Distribution in Rectangular Channels Using Entropy Concept
This study makes use of the Tsallis entropy to predict the shear stress distribution. Given a definition of the Tsallis entropy, it is maximized by using the probability density function, which then is used to attain a novel shear stress equation. This is then employed for calculating the shear stress distribution in rectangular channels in different aspect ratios and finally, for viability, th...
متن کامل