Bio-Inspired and Information-Theoretic Signal Processing

2015 ◽  
pp. 493-532
Author(s):  
Romis de Faissol ◽  
Levy Boccato ◽  
Denis Fantinato ◽  
Jugurta Filho ◽  
Aline de Oliveira ◽  
...  
Keyword(s):  
Signal Processing ◽  
Information Theoretic

scholarly journals Data Discovery and Anomaly Detection Using Atypicality for Real-Valued Data

Entropy ◽  
2019 ◽  
Vol 21 (3) ◽  
pp. 219 ◽  
Author(s):  
Elyas Sabeti ◽  
Anders Høst-Madsen
Keyword(s):  
Heart Rate ◽  
Signal Processing ◽  
Heart Rate Variability ◽  
Big Data ◽  
Anomaly Detection ◽  
Minimum Description Length ◽  
Data Discovery ◽  
Information Theoretic ◽  
Typical Data ◽  
Insight Into

The aim of using atypicality is to extract small, rare, unusual and interesting pieces out of big data. This complements statistics about typical data to give insight into data. In order to find such “interesting” parts of data, universal approaches are required, since it is not known in advance what we are looking for. We therefore base the atypicality criterion on codelength. In a prior paper we developed the methodology for discrete-valued data, and the current paper extends this to real-valued data. This is done by using minimum description length (MDL). We develop the information-theoretic methodology for a number of “universal” signal processing models, and finally apply them to recorded hydrophone data and heart rate variability (HRV) signal.

Closed-Form Expressions of Some Stochastic Adapting Equations for Nonlinear Adaptive Activation Function Neurons

2003 ◽  
Vol 15 (12) ◽  
pp. 2909-2929 ◽  
Author(s):  
Simone Fiori
Keyword(s):  
Differential Equation ◽  
Signal Processing ◽  
Blind Deconvolution ◽  
Activation Function ◽  
Activation Functions ◽  
Neuron Models ◽  
Artificial Neuron ◽  
Information Theoretic ◽  
Blind Signal Processing ◽  
Blind Signal

In recent work, we introduced nonlinear adaptive activation function (FAN) artificial neuron models, which learn their activation functions in an unsupervised way by information-theoretic adapting rules. We also applied networks of these neurons to some blind signal processing problems, such as independent component analysis and blind deconvolution. The aim of this letter is to study some fundamental aspects of FAN units' learning by investigating the properties of the associated learning differential equation systems.

On the Detection of the Number of Signals in Multilead ECGs

1992 ◽  
Vol 31 (04) ◽  
pp. 247-255 ◽  
Author(s):  
A. van Oosterom ◽  
G. J. H. Uijen
Keyword(s):  
Signal Processing ◽  
Covariance Matrix ◽  
Body Surface ◽  
Threshold Method ◽  
Multiple Signals ◽  
Information Theoretic ◽  
New Methods ◽  
Surface Maps ◽  
Information Theoretic Criteria ◽  
Source Signals

Abstract:New methods which have been developed in the field of signal processing to detect the number of intrinsic source signals in a set of multiple signals have been studied in their application to the analysis of multilead ECGs (body surface maps). All these methods are based on the analysis of the eigenvalues of the covariance matrix related to the observed data. By using artificial multilead ECGs in which the properties of the signal and noise parts were known, the effect of taking into account these properties was investigated. It was found that the method based on information-theoretic criteria was superior in detecting the number of signal components, in the case that the noise in the data is non-white, as compared to a simple threshold method.

Sign in / Sign up

Export Citation Format

Share Document