Results 21 to 30 of about 580,652 (288)
Optimising Distributions with Natural Gradient Surrogates
CoRR, 2023 Natural gradient methods have been used to optimise the parameters of probability distributions in a variety of settings, often resulting in fast-converging procedures. Unfortunately, for many distributions of interest, computing the natural gradient has
Turner, Richard E., So, Jonathan
core +3 more sources
On the Natural Gradient of the Evidence Lower Bound
J. Mach. Learn. Res., 2023 This article studies the Fisher-Rao gradient, also referred to as the natural gradient, of the evidence lower bound (ELBO) which plays a central role in generative machine learning. It reveals that the gap between the evidence and its lower bound, the ELBO, has essentially a vanishing natural gradient within unconstrained optimization.
Ay, Nihat +2 more
openaire +3 more sources
Natural gradients for deformable registration [PDF]
2010 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2010 We apply the concept of natural gradients to deformable registration. The motivation stems from the lack of physical interpretation for gradients of image-based difference measures. The main idea is to endow the space of deformations with a distance metric which reflects the variation of the difference measure between two deformations.
Darko Zikic, Ali Kamen, Nassir Navab
openaire +1 more source
Sparse Channel Estimation for Interference Limited OFDM Systems and Its Convergence Analysis
IEEE Access, 2017 Wireless communication channels are highly prone to interference in addition to the presence of additive white Gaussian noise (AWGN). Stochastic gradient (SG)-based non-parametric maximum likelihood (NPML) estimator, gives better channel estimates in the
Abhijeet Bishnu, Vimal Bhatia
doaj +1 more source
An approach to nonlinear viscoelasticity via metric gradient flows [PDF]
, 2013 We formulate quasistatic nonlinear finite-strain viscoelasticity of rate type as a gradient system. Our focus is on nonlinear dissipation functionals and distances that are related to metrics on weak diffeomorphisms and that ensure time-dependent frame ...
Mielke, A. +4 more
core +1 more source
Natural gradient matrix momentum [PDF]
, 1999 Natural gradient learning is an efficient and principled method for improving online learning. In practical applications there will be an increased cost required in estimating and inverting the Fisher information matrix.
Saad David +2 more
core +2 more sources
On Performance and Calibration of Natural Gradient Langevin Dynamics
IEEE Access, 2023 Producing deep neural network (DNN) models with calibrated confidence is essential for applications in many fields, such as medical image analysis, natural language processing, and robotics.
Hanif Amal Robbani +3 more
doaj +1 more source
Ecological Indicators, 2021 The functional response of natural patches to surrounding land-use changes is strongly related to variations in functional traits of coexisting species.
Erika Bazzato +3 more
doaj +1 more source
Research progress in biomimetic gradient scaffolds for osteochondral tissue engineering
Cailiao gongcheng, 2022 Osteochondral defects are the main cause of joint morbidity and disability in elderly patients, and osteochondral tissue engineering is one of the methods to repair osteochondral defects.
WAN Li +5 more
doaj +1 more source
On the Fisher Metric of Conditional Probability Polytopes
Entropy, 2014 We consider three different approaches to define natural Riemannian metrics on polytopes of stochastic matrices. First, we define a natural class of stochastic maps between these polytopes and give a metric characterization of Chentsov type in terms of ...
Guido Montúfar, Johannes Rauh, Nihat Ay
doaj +1 more source