Results 61 to 70 of about 6,455,829 (284)
Natural Gradient Flow in the Mixture Geometry of a Discrete Exponential Family
Entropy, 2015 In this paper, we study Amari’s natural gradient flows of real functions defined on the densities belonging to an exponential family on a finite sample space.
Luigi Malagò, Giovanni Pistone
doaj +1 more source
Natural Conjugate Gradient in Variational Inference [PDF]
, 2008 Variational methods for approximate inference in machine learning often adapt a parametric probability distribution to optimize a given objective function. This view is especially useful when applying variational Bayes (VB) to models outside the conjugate-exponential family. For them, variational Bayesian expectation maximization (VB EM) algorithms are
Antti Honkela +3 more
openaire +2 more sources
Stochastic search using the natural gradient [PDF]
Proceedings of the 26th Annual International Conference on Machine Learning, 2009 To optimize unknown 'fitness' functions, we present Natural Evolution Strategies, a novel algorithm that constitutes a principled alternative to standard stochastic search methods. It maintains a multinormal distribution on the set of solution candidates.
Yi Sun 0003 +3 more
openaire +2 more sources
Metastatic niche shaped by host factors influences disseminated cancer cell fate
FEBS Letters, EarlyView.Metastasis is shaped not only by cancer cells but also by the environments they encounter. This review explores how factors such as aging, diet, the microbiome, lifestyle, and environmental exposures remodel organ‐specific niches in the lung, liver, bone, and brain, influencing where metastatic cells survive, remain dormant, or grow, and ultimately ...
Gwennan Delyth Ward +2 more
wiley +1 more source
Affine Calculus for Constrained Minima of the Kullback–Leibler Divergence
StatsThe non-parametric version of Amari’s dually affine Information Geometry provides a practical calculus to perform computations of interest in statistical machine learning.
Giovanni Pistone
doaj +1 more source
Functional Natural Policy Gradients
CoRRWe propose a cross-fitted debiasing device for policy learning from offline data. A key consequence of the resulting learning principle is $\sqrt N$ regret even for policy classes with complexity greater than Donsker, provided a product-of-errors nuisance remainder is $O(N^{-1/2})$.
Aurélien F. Bibaut +3 more
openaire +3 more sources
Going viral − Virus‐targeting nanobodies for research, diagnostics and therapy
FEBS Letters, EarlyView.The increasing prevalence of viral diseases represents a major threat for global human health. Nanobodies, small binding fragments derived from camelid heavy‐chain antibodies, offer versatile tools for viral research, diagnostics, and antiviral therapy.
Domenic Ebert, Ulrich Rothbauer
wiley +1 more source
Mirror Descent and Exponentiated Gradient Algorithms Using Trace-Form Entropies
EntropyThis paper introduces a broad class of Mirror Descent (MD) and Generalized Exponentiated Gradient (GEG) algorithms derived from trace-form entropies defined via deformed logarithms. Leveraging these generalized entropies yields MD and GEG algorithms with
Andrzej Cichocki +3 more
doaj +1 more source
Vadose Zone Journal, 2022 Clay content in soil potentially has strong influence on soil structure and thereby controls soil physical, hydraulic, and gas transport properties. Only few studies have investigated the effect of clay content on saturated hydraulic conductivity (Ks ...
Ting Yang +5 more
doaj +1 more source
Comparing the paths of Hedgehog release in artificial and natural settings
FEBS Letters, EarlyView.Hh morphogens require solubilization factors to overcome their tight association with the plasma membrane and signal to distant cells, but the precise functions of these factors are controversial. Here, we use published in vivo observations as the most relevant evidence to test their suggested functions and discuss contradictions and gaps between in ...
Janna Puschmann, Kay Grobe
wiley +1 more source