Results 21 to 30 of about 2,630,587 (176)

Adversarial Robustness via Fisher-Rao Regularization [PDF]

, 2022
Adversarial robustness has become a topic of growing interest in machine learning since it was observed that neural networks tend to be brittle. We propose an information-geometric formulation of adversarial defense and introduce FIRE, a new Fisher-Rao ...
Labeau, Fabrice, Ayed, Ismail Ben, Picot, Marine, Boudiaf, Malik, Messina, Francisco, Piantanida, Pablo, Fabrice Labeau, Pablo Piantanida, Malik Boudiaf, Ismail Ben Ayed, Marine Picot, Francisco Messina   +11 more
core   +1 more source

Extending the Extreme Physical Information to Universal Cognitive Models via a Confident Information First Principle

Entropy, 2014
The principle of extreme physical information (EPI) can be used to derive many known laws and distributions in theoretical physics by extremizing the physical information loss K, i.e., the difference between the observed Fisher information I and the ...
Xiaozhao Zhao, Yuexian Hou, Dawei Song, Wenjie Li   +3 more
doaj   +1 more source

Weibull-Modeled Impulsive Noise in Wireless Communication Systems: An Information-Geometric Approach

IEEE Open Journal of the Communications Society
Automatic modulation classification (AMC) enables cognitive wireless receivers to identify modulation formats under dynamic channel conditions. However, impulsive interference severely degrades accuracy when receiver preprocessing and neural network ...
Mehdi Rahimi, Samaneh Bidabadi, Mehdi Korki, Felipe A. P. de Figueiredo   +3 more
doaj   +1 more source

Cramér-Rao Bound Study of Multiple Scattering Effects in Target Separation Estimation

International Journal of Antennas and Propagation, 2013
The information about the distance of separation between two-point targets that is contained in scattering data is explored in the context of the scalar Helmholtz operator via the Fisher information and associated Cramér-Rao bound (CRB) relevant to ...
Edwin A. Marengo, Paul Berestesky
doaj   +1 more source

On Closed-Form Expressions for the Fisher-Rao Distance [PDF]

The Fisher-Rao distance is the geodesic distance between probability distributions in a statistical manifold equipped with the Fisher metric, which is a natural choice of Riemannian metric on such manifolds. It has recently been applied to supervised and
Meneghetti, Fábio C. C., Pinele, Julianna, Miyamoto, Henrique K., Costa, Sueli I. R.   +3 more
core   +1 more source

Intrinsic Losses Based on Information Geometry and Their Applications

Entropy, 2017
One main interest of information geometry is to study the properties of statistical models that do not depend on the coordinate systems or model parametrization; thus, it may serve as an analytic tool for intrinsic inference in statistics. In this paper,
Yao Rong, Mengjiao Tang, Jie Zhou
doaj   +1 more source

A shape distance based on the Fisher–Rao metric and its application for shapes clustering

, 2017
In the clustering of shapes, which is a longstanding challenge in the framework of geometric morphometrics and shape analysis, is crucial the selection and application of a suitable and appropriate measurement of distance among observations (i.e ...
De Sanctis, A, De Sanctis A., Stefano Antonio Gattone, DE SANCTIS, Angela Anna, Pulcini, D, Tommaso Russo, Russo, T, Gattone S. A., Pulcini D., Angela De Sanctis, Domitilla Pulcini, Russo T., GATTONE, Stefano Antonio, Gattone, Sa, Pulcini, Domitilla, Russo, Tommaso   +15 more
core   +1 more source

The Fisher-Rao metric

, 2008
The information in a mass of data is summarised by fitting to the data a probability density function (pdf) chosen from a parameterised family of pdfs. Distances between pairs of pdfs are calculated using the Fisher-Rao metric, which provides a measure ...
Maybank, Stephen J.
core   +4 more sources

BioTraj: Information-Geometric Mapping of Spatial and Temporal Heterogeneity in Biological Trajectories

Mathematics
Quantitative trajectory analysis commonly summarizes biological motion using global geometric or dynamical descriptors, but these do not directly quantify how movement statistics vary across space or time.
Danilo Roccatano
doaj   +1 more source

Sensor Networks for Optimal Target Localization with Bearings-Only Measurements in Constrained Three-Dimensional Scenarios

Sensors, 2013
In this paper, we address the problem of determining the optimal geometric configuration of an acoustic sensor network that will maximize the angle-related information available for underwater target positioning.
Joaquin Aranda, David Moreno-Salinas, Antonio Pascoal   +2 more
doaj   +1 more source