The Approximate Divergence Operator

t-divergence Based Approximate Inference

Approximate inference is an important technique for dealing with large, intractable graphical models based on the exponential family of distributions. We extend the idea of approximate inference to the t-exponential family by defining a new t-divergence. This divergence measure is obtained via convex duality between the log-partition function of the t-exponential family and a new t-entropy. We ...

Bregman divergence as relative operator entropy

vet C ˜e — ™onvex set in — f—n—™h sp—™e —nd r ˜e — ril˜ert sp—™eF por —n oper—torE v—lued smooth fun™tion—l © X C 3 B@rAD we dene the Bregman divergence of x; y P C —s D © @x; yA Xa ©@xA ©@yA lim t3CH t I ©@y C t@x yAA ©@yA : ‚e™—ll th—t origin—lly vev fregm—n introdu™ed this ™on™ept for re—lEv—lued ™onvex fun™E tions © ‘P“F yn the selfE—djoint oper—torsD there is — st—nd—rd p—rti—l orderingD A...

Approximate analytical formulas for Kirchhoff migration operator

Two important tools in seismic processing––Kirchhoff migration and demigration operators––are the basis for many imaging problems solution. Due to high numerical and computational requirements, the use of those tools for three dimensions are very computation-costly. This fact has motivated us to investigate Kirchhoff migration operations for simpler types of media in order to provide faster res...

A Quasi-Local Interpolation Operator Preserving the Discrete Divergence

An Approximate Inverse to the Extended Born Modeling Operator

We modify RTM to create an approximate inverse to the extended Born modeling operator in 2D. The derivation uses asymptotic ray theory and stationary phase principle, but the result applies directly to RTM. The inverse operator differs from the adjoint operator only by application of several explicit velocity-independent filters. This inverse operator, on the one hand, can be used as true ampli...