Search results for "NEURAL NETWORKS"
showing 10 items of 599 documents
Ghost stochastic resonance in FitzHugh–Nagumo circuit
2014
International audience; The response of a neural circuit submitted to a bi-chromatic stimulus and corrupted by noise is investigated. In the presence of noise, when the spike firing of the circuit is analysed, a frequency not present at the circuit input appears. For a given range of noise intensities, it is shown that this ghost frequency is almost exclusively present in the interspike interval distribution. This phenomenon is for the first time shown experimentally in a FitzHugh-Nagumo circuit.
Neural Classification of Compost Maturity by Means of the Self-Organising Feature Map Artificial Neural Network and Learning Vector Quantization Algo…
2019
Self-Organising Feature Map (SOFM) neural models and the Learning Vector Quantization (LVQ) algorithm were used to produce a classifier identifying the quality classes of compost, according to the degree of its maturation within a period of time recorded in digital images. Digital images of compost at different stages of maturation were taken in a laboratory. They were used to generate an SOFM neural topological map with centres of concentration of the classified cases. The radial neurons on the map were adequately labelled to represent five suggested quality classes describing the degree of maturation of the composted organic matter. This enabled the creation of a neural separator classify…
Theoretical and experimental study of two discrete coupled Nagumo chains
2001
We analyze front wave (kink and antikink) propagation and pattern formation in a system composed of two coupled discrete Nagumo chains using analytical and numerical methods. In the case of homogeneous interaction among the chains, we show the possibility of the effective control on wave propagation. In addition, physical experiments on electrical chains confirm all theoretical behaviors.
Seizure Prediction Using EEG Channel Selection Method
2022
Seizure prediction using intracranial electroencephalogram (iEEG) is still challenging because of complicated signals in spatial and time domains. Feature selection in the spatial domain (i.e., channel selection) has been largely ignored in this field. Hence, in this paper, a novel approach of iEEG channel selection strategy combined with one-dimensional convolutional neural networks (1D-CNN) was presented for seizure prediction. First, 15-sec and 30-sec iEEG segments with an increasing number of channels (from one channel to all channels) were sequentially fed into 1D-CNN models for training and testing. Then, the channel case with the best classification rate was selected for each partici…
Estimating Programming Exercise Difficulty using Performance Factors Analysis
2020
This Work in Progress Paper studies student and exercise modelling based on pass/fail log data gathered from an introductory programming course. Contemporary education capitalizes on the communications technology and remote study. This can create distance between the teacher and students and the resulting lack of awareness of the difficulties students encounter can lead to low student satisfaction, dropout and poor grades. In many cases, various technological solutions are used to collect individual exercise submissions, but there are little resources for indexing or modelling the exercises in depth. Exercise specific feedback from students may not be easily obtainable either. In the presen…
PROPAGATING INTERFACES IN A TWO-LAYER BISTABLE NEURAL NETWORK
2006
The dynamics of propagating interfaces in a bistable neural network is investigated. We consider the network composed of two coupled 1D lattices and assume that they interact in a local spatial point (pin contact). The network unit is modeled by the FitzHugh–Nagumo-like system in a bistable oscillator mode. The interfaces describe the transition of the network units from the rest (unexcited) state to the excited state where each unit exhibits periodic sequences of excitation pulses or action potentials. We show how the localized inter-layer interaction provides an "excitatory" or "inhibitory" action to the oscillatory activity. In particular, we describe the interface propagation failure a…
Pinning of a kink in a nonlinear diffusive medium with a geometrical bifurcation: Theory and experiments
2004
International audience; We study the dynamics of a kink propagating in a Nagumo chain presenting a geometrical bifurcation. In the case of weak couplings, we define analytically and numerically the coupling conditions leading to the pinning of the kink at the bifurcation site. Moreover, real experiments using a nonlinear electrical lattice confirm the theoretical and numerical predictions.
Estudio de la radiación neta en zonas semiáridas utilizando modelos lineales y neuronales y la sinergia entre GERB y SEVIRI
2012
Las regiones áridas o semiáridas se caracterizan por una distribución irregular de los recursos hídricos, lo que muchas veces constituye una limitación para el desarrollo de una determinada región. La variabilidad hidrológica de estas regiones se debe a la mala distribución espacial y temporal de la lluvia, a la topografía heterogénea y a los cambios de origen antropogénicos que muchas veces conducen a procesos de degradación y de desertificación. Como en estas regiones la evapotranspiración explica una parte significativa de la pérdida de agua hacia la atmósfera, el estudio y modelización de la radiación neta en superficie (Rn), es de suma importancia, una vez que las estimaciones o medici…
Contributions and applications around low resource deep learning modeling
2023
El aprendizaje profundo representa la vanguardia del aprendizaje automático en multitud de aplicaciones. Muchas de estas tareas requieren una gran cantidad de recursos computacionales, lo que limita su adopción en dispositivos integrados. El objetivo principal de esta tesis es estudiar métodos y algoritmos que permiten abordar problemas utilizando aprendizaje profundo con bajos recursos computacionales. Este trabajo también tiene como objetivo presentar aplicaciones de aprendizaje profundo en la industria. La primera contribución es una nueva función de activación para redes de aprendizaje profundo: la función de módulo. Los experimentos muestran que la función de activación propuesta logra…