ON THE LOCAL-FIELD DISTRIBUTION IN ATTRACTOR NEURAL NETWORKS

1996 ◽  
Vol 07 (04) ◽  
pp. 463-483 ◽  
Author(s):  
E. KORUTCHEVA ◽  
K. KOROUTCHEV
Keyword(s):  
Neural Networks ◽  
Recurrent Neural Networks ◽  
Dynamic Properties ◽  
Basin Of Attraction ◽  
Mean Field ◽  
Field Approximation ◽  
Local Fields ◽  
Mean Field Approximation ◽  
The Mean ◽  
The Basin Of Attraction

In this paper a simple two-layer neural network's model, similar to that studied by D. Amit and N. Brunel,11 is investigated in the frames of the mean-field approximation. The distributions of the local fields are analytically derived and compared to those obtained in Ref. 11. The dynamic properties are discussed and the basin of attraction in some parametric space is found. A procedure for driving the system into a basin of attraction by using a regulation imposed on the network is proposed. The effect of outer stimulus is shown to have a destructive influence on the attractor, forcing the latter to disappear if the distribution of the stimulus has high enough variance or if the stimulus has a spatial structure with sufficient contrast. The techniques, used in this paper, for obtaining the analytical results can be applied to more complex topologies of linked recurrent neural networks.

Attractor Dynamics in Feedforward Neural Networks

2000 ◽  
Vol 12 (6) ◽  
pp. 1313-1335 ◽  
Author(s):  
Lawrence K. Saul ◽  
Michael I. Jordan
Keyword(s):  
Neural Networks ◽  
Mean Field ◽  
Feedforward Neural Networks ◽  
Field Approximation ◽  
Generative Models ◽  
Mean Field Approximation ◽  
Field Equations ◽  
Attractor Dynamics ◽  
Mean Field Equations ◽  
The Mean

We study the probabilistic generative models parameterized by feedfor-ward neural networks. An attractor dynamics for probabilistic inference in these models is derived from a mean field approximation for large, layered sigmoidal networks. Fixed points of the dynamics correspond to solutions of the mean field equations, which relate the statistics of each unittothoseofits Markovblanket. We establish global convergence of the dynamics by providing a Lyapunov function and show that the dynamics generate the signals required for unsupervised learning. Our results for feedforward networks provide a counterpart to those of Cohen-Grossberg and Hopfield for symmetric networks.

scholarly journals The Combined Effect of Rotation and Magnetic Field on Finite-Amplitude Thermal Convection

1973 ◽  
Vol 26 (5) ◽  
pp. 617 ◽  
Author(s):  
R Van der Borght ◽  
JO Murphy
Keyword(s):  
Magnetic Field ◽  
Mean Field ◽  
Field Approximation ◽  
Finite Amplitude ◽  
Combined Effect ◽  
Mean Field Approximation ◽  
Free Boundaries ◽  
Wide Range ◽  
The Mean ◽  
Basic Equations

The combined effect of an imposed rotation and magnetic field on convective transfer in a horizontal Boussinesq layer of fluid heated from below is studied in the mean field approximation. The basic equations are derived by a variational technique and their solutions are then found over a wide range of conditions, in the case of free boundaries, by numerical and analytic techniques, in particular by asymptotic and perturbation methods. The results obtained by the different techniques are shown to be in excellent agreement. As for the linear theory, the calculations predict that the simultaneous presence' of a magnetic field and rotation may produce conflicting tendencies.

scholarly journals Locating the critical end point using the linear sigma model coupled to quarks.

2018 ◽  
Vol 172 ◽  
pp. 02003
Author(s):  
Alejandro Ayala ◽  
J. A. Flores ◽  
L. A. Hernández ◽  
S. Hernández-Ortiz
Keyword(s):  
Sigma Model ◽  
Chemical Potential ◽  
Potential Temperature ◽  
Mean Field ◽  
Field Approximation ◽  
Baryon Density ◽  
Linear Sigma Model ◽  
Mean Field Approximation ◽  
End Point ◽  
The Mean

We use the linear sigma model coupled to quarks to compute the effective potential beyond the mean field approximation, including the contribution of the ring diagrams at finite temperature and baryon density. We determine the model couplings and use them to study the phase diagram in the baryon chemical potential-temperature plane and to locate the Critical End Point.

scholarly journals UNDERSTANDING MAGNETIC INSTABILITY IN GAPLESS SUPERCONDUCTORS

2006 ◽  
Vol 21 (04) ◽  
pp. 910-913 ◽  
Author(s):  
Mei Huang
Keyword(s):  
Superconducting State ◽  
Phase Fluctuation ◽  
Mean Field ◽  
Field Approximation ◽  
Mean Field Approximation ◽  
Bcs Theory ◽  
Magnetic Instability ◽  
Fermi Surfaces ◽  
Strong Phase ◽  
The Mean

Magnetic instability in gapless superconductors still remains as a puzzle. In this article, we point out that the instability might be caused by using BCS theory in mean-field approximation, where the phase fluctuation has been neglected. The mean-field BCS theory describes very well the strongly coherent or rigid superconducting state. With the increase of mismatch between the Fermi surfaces of pairing fermions, the phase fluctuation plays more and more important role, and "soften" the superconductor. The strong phase fluctuation will eventually quantum disorder the superconducting state, and turn the system into a phase-decoherent pseudogap state.

Quartetting in Fermionic Matter and Alpha-Particle Condensation in Nuclear Systems

2006 ◽  
Vol 21 (31n33) ◽  
pp. 2513-2546 ◽  
Author(s):  
G. Röpke ◽  
P. Schuck
Keyword(s):  
Nuclear Matter ◽  
Cluster Formation ◽  
Mean Field ◽  
Field Approximation ◽  
Mean Field Approximation ◽  
Density Region ◽  
Nuclear Matter Density ◽  
Nuclear Systems ◽  
The Mean ◽  
Finite Nuclei

Quantum condensates in nuclear matter are treated beyond the mean-field approximation, with the inclusion of cluster formation. The occurrence of a separate binding pole in the four-particle propagator in nuclear matter is investigated with respect to the formation of a condensate of α-like particles (quartetting), which is dependent on temperature and density. Due to Pauli blocking, the formation of an α-like condensate is limited to the low-density region. Consequences for finite nuclei are considered. In particular, excitations of self-conjugate 2n-Z–2n-N nuclei near the n-α-breakup threshold are candidates for quartetting. We review some results and discuss their consequences. Exploratory calculations are performed for the density dependence of the α condensate fraction at zero temperature to address the suppression of the four-particle condensate below nuclear-matter density.

Sign in / Sign up

Export Citation Format

Share Document