About symmetric rank-1 random matrices

dg.differential geometry – Projections in boundless dimensional statistical manifolds Answer

Hello pricey customer to our community We will proffer you an answer to this query dg.differential geometry – Projections in boundless dimensional statistical manifolds ,and the respond will breathe typical by documented info sources, We welcome you and proffer you fresh questions and solutions, Many customer are questioning in regards to the respond to this query.

dg.differential geometry – Projections in boundless dimensional statistical manifolds

Are you attempting to assemble an infinite-dimensional (Hilbert) manifold of likelihood measures on a set measurable area? Or you merely need to discover a Bergman divergence analog in organize to generalize a divergence-like metric in an infinite-dimensional manifold? It just isn’t completely limpid within the OP what kindly of analog you might be searching for.

In a finite-dimensional statistical manifold, for instance an exponential household with unaffected parameterizations, the manifold does include a canonical connection as you talked about. In the finite-dimensional status, since we are able to take a parameterization, the geometry of the statistical manifold is embedded within the parameter area. The picture of such a projection to a $nabla$-flat submanifold would coincide to a sub-family of the exponential household, too a restricted assortment of parameters that outline this sub-family.

In an infinite-dimensional statistical manifold, assuming that you just need to examine two factors; you’ll be able to merely compute the norms of their dissimilarity and different kindly of similarity measures are workable [Harandi et.al.]. If you might be speaking about infinite-dimensional statistical manifold that consists of likelihood measures that can’t breathe The inside product construction love we’ve for low-dimensional parameter area, nevertheless, doesn’t at all times live. In a particular illustration identified by [Newton2], we are able to employ an $alpha$-divergence (which is a generalization to Bergman divergence, behold this submit from stat.SE) to partially narrate an identical inside product-like construction for an infinite-dimensional statistical manifold. The asymptotic outcomes from geometric perspective [Kass&Vos] could generalize into infinite-dimensional status if you’ll find a “good finite-dimensional basis approximation”. However, a strict inside product construction just isn’t at all times workable in infinite-dimensional status, due to this fact projections could not generalize. If it does generalize, its acceptation would doubtless breathe a sub-family of likelihood measures.

In one other course, if you happen to want to assemble a realization (or haul samples) from an inifinite-dimensional statistical manifold, the place to begin would breathe Dirichlet processes wiki. And [Newton] proposed to employ transformation (Fenchel–Legendre rework) strategy to assemble fresh infinite-dimensional statistical manifolds in a extra summary route.

Reference

[Newton2] Infinite-dimensional statistical manifolds based mostly on a balanced chart, 2016.

[Harandi et.al.] Bregman Divergences for Infinite Dimensional Covariance Matrices,2014.

[Newton] An infinite-dimensional statistical manifold modelled on Hilbert area,2012.

[Kass&Vos] Geometrical Foundations of Asymptotic Inference, 1997.

we are going to proffer you the answer to dg.differential geometry – Projections in boundless dimensional statistical manifolds query by way of our community which brings all of the solutions from a number of dependable sources.

Add comment