Results 41 to 50 of about 296,906 (157)

Using rényi-divergence and arimoto-rényi information to quantify membership information leakage [PDF]

open access: yes, 2021
Membership inference attacks, i.e., adversarial attacks inferring whether a data record is used for training a machine learning model, has been recently shown to pose a legitimate privacy risk in machine learning literature. In this paper, we propose two
Farokhi, F
core   +1 more source

Variational Inference via Rényi Bound Optimization and Multiple-Source Adaptation

open access: yesEntropy, 2023
Variational inference provides a way to approximate probability densities through optimization. It does so by optimizing an upper or a lower bound of the likelihood of the observed data (the evidence).
Dana Zalman (Oshri), Shai Fine
doaj   +1 more source

Simultaneous Network Design With Restricted Link Usage

open access: yesNetworks, EarlyView.
ABSTRACT Given a digraph with two terminal vertices s$$ s $$ and t$$ t $$ as well as a conservative cost function and several not necessarily disjoint color classes on its arc set, our goal is to find a minimum‐cost subset of the arcs such that its intersection with each color class contains an s$$ s $$‐t$$ t $$ dipath.
Naonori Kakimura   +3 more
wiley   +1 more source

Rényi Divergence and Kullback-Leibler Divergence

open access: yes, 2013
Accepted by IEEE Transactions on Information Theory. To appear.Rényi divergence is related to Rényi entropy much like Kullback-Leibler divergence is related to Shannon's entropy, and comes up in many settings.
Harremoës, Peter, van Erven, Tim
core   +1 more source

Point Divergence Gain and Multidimensional Data Sequences Analysis

open access: yesEntropy, 2018
We introduce novel information-entropic variables—a Point Divergence Gain ( Ω α ( l → m ) ), a Point Divergence Gain Entropy ( I α ), and a Point Divergence Gain Entropy Density ( P α )—which are derived from the Rényi entropy and ...
Renata Rychtáriková   +3 more
doaj   +1 more source

Rényi Entropy of k-Records

open access: yesRevstat Statistical Journal, 2022
In this paper, we discuss on Rényi entropy of k-records arising from any continuous distribution in detail. The relevance of constructing k-records from random sample in the context of information contained in a random variable has been described in the
Jitto Jose , E. I. Abdul Sathar
doaj   +1 more source

Vendi information gain for active learning and its application to ecology

open access: yesMethods in Ecology and Evolution, EarlyView.
Abstract While monitoring biodiversity through camera traps has become an important endeavour for ecological research, identifying species in the captured image data remains a major bottleneck due to limited labelling resources. Active learning—a machine learning paradigm that selects the most informative data to label and train a predictive model ...
Quan Nguyen, Adji Bousso Dieng
wiley   +1 more source

Hardness of (Semiuniform) MLWE with Short Distributions Using the Rényi Divergence

open access: yesIET Information Security, 2023
The module learning with errors (MLWE) problem has attracted considerable attention for its tradeoff between security and efficiency. The quantum/classical worst-case to average-case hardness for the MLWE problem (or more exactly, a family of problems ...
Wenjuan Jia, Baocang Wang
doaj   +1 more source

Minimizing cutting costs in 1D rod cutting

open access: yesInternational Transactions in Operational Research, EarlyView.
Abstract We study a one‐dimensional rod‐cutting problem arising from an industrial setting where cutting itself carries cost. Each order specifies a length interval, and the task is to assign orders to warehouse rods so that all orders are satisfied while the number of cuts is minimized.
Bowen Li, Attila Sali
wiley   +1 more source

An active contour model for texture image segmentation using Rényi divergence measure

open access: yesMathematical Modelling and Analysis, 2022
This paper proposes an efficient method for active unsupervised texture segmentation. A new descriptor for texture features extractions based on Gaussian and mean curvature is constructed.
Sidi Yassine Idrissi
doaj   +1 more source

Home - About - Disclaimer - Privacy