Search

The Blume-Emery-Griffiths model with a random crystal field is studied in a random graph architecture, in which the average connectivity is a controllable parameter. The disordered average over the graph realizations is treated by replica symmetry formalism of order parameter functions. A self-consistent equation for the distribution of local fields is derived and numerically solved by a population dynamics algorithm. The results show that the average connectivity amounts to changes in the topology of the phase diagrams. Phase diagrams for representative values of the model parameters are compared with those obtained for fully connected mean field and renormalization group approaches., Comment: 14 pages, 6 figures, accepted for publication in Physica A

High lipid and high protein infant formula milk powders were stored at water activity of 0.11, 0.33 and 0.53 for up to fourteen weeks at 40 °C to investigate the effect of storage water activity on physicochemical properties and formation of volatiles to thereby recommend optimal storage water activity conditions. Water activity of the powders was determined during storage together with surface colour, glass transition temperature combined with dynamic headspace sampling followed by gas chromatography/mass spectrometry. The principal component analysis (PCA) showed that the optimal water activity for storage of high lipid infant formula milk powder, for which lipid oxidation was found to be the critical quality parameter, is aw = 0.33 with lowest lipid oxidation, while for high protein infant formula milk powder, for which protein degradation was found to be the critical quality parameter, aw = 0.11 is optimal to limit formation of Maillard reaction products.

The plastic character of brain synapses is considered to be one of the foundations for the formation of memories. There are numerous kinds of such phenomenon currently described in the literature, but their role in the development of information pathways in neural networks with recurrent architectures is still not completely clear. In this paper we study the role of an activity-based process, called pre-synaptic dependent homeostatic scaling, in the organization of networks that yield precise-timed spiking patterns. It encodes spatio-temporal information in the synaptic weights as it associates a learned input with a specific response. We introduce a correlation measure to evaluate the precision of the spiking patterns and explore the effects of different inhibitory interactions and learning parameters. We find that large learning periods are important in order to improve the network learning capacity and discuss this ability in the presence of distinct inhibitory currents.

The ability of a metric attractor neural networks (MANN) to learn structured patterns is analyzed. In particular we consider collections of fingerprints, which present some local features, rather than being modeled by random patterns. The network retrieval proved to be robust to varying the pattern activity, the threshold strategy, the topological arrangement of the connections, and for several types of noisy configuration. We found that the lower the fingerprint patterns activity is, the higher the load ratio and retrieval quality are. A simplified theoretical framework, for the unbiased case, is developed as a function of five parameters: the load ratio, the finiteness connectivity, the density degree of the network, randomness ratio, and the spatial pattern correlation. Linked to the latter appears a new neural dynamics variable: the spatial neural correlation. The theory agrees quite well with the experimental results.

In this paper we investigate the three-state spins van Hemmen model with a crystalline field in the random network. The van Hemmen model is known for being treated without the use of replicas. The same method used by van Hemmen is utilized here to average over disordered exchange interactions. To deal with averages over the realizations of the lattice we utilized the replica symmetry formalism of order parameter functions. The order parameters calculated to detect which phase the system encounters are calculated numerically by means of a population dynamics algorithm. Firstly we obtained phase diagrams in low fixed temperature in the diagram crystalline field versus random field, we observe higher connectivities to give rise to segregation between high activity and low activity spin glass (SG) phases, also we verify the appearance of a tricritical point in the low crystalline field region. Finally to account the effects of high thermal fluctuations we have drawn phase diagrams in the temperature versus random field plane for fixed values of crystalline field. Here we observe an important modification of the phase plane topology by the increment of network connectivity. It is also notable the presence of a reentrant behavior in first and second order phase transitions when the crystalline field is sufficiently large.

The Random Field Ising Model (RFIM) following bimodal and Gaussian distributions for the RF is investigated using a finite connectivity technique. We focused on determining the order of the phase transition as well as the existence of a tricritical point as a function of the connectivity c for both types of RF distribution. Our results indicate that for the Gaussian distribution the phase transition is always second-order. For the bimodal distribution, there is indeed a tricritical point. However, its location is strongly dependent on c . The tricritical point is suppressed below a certain minimum value of connectivity.

A synaptic connectivity model is assembled on a spiking neuron network aiming to build up a dynamic pattern recognition system. The connection architecture includes gap junctions and both inhibitory and excitatory chemical synapses based on Hebb’s hypothesis. The network evolution resulting from external stimulus is sampled in a properly defined frequency space. Neurons’ responses to different current injections are mapped onto a subspace using Principal Component Analysis. Departing from the base attractor, related to a quiescent state, different external stimuli drive the network to different fixed points through specific trajectories in this subspace.

Sparse-code networks have retrieval abilities which are strongly dependent on the firing threshold for the neurons. If the connections are spatially uniform, the macroscopic properties of the network can be measured by the overlap between neurons and learned patterns, and by the global activity. However, for nonuniform networks, for instance small-world networks, the neurons can retrieve fragments of patterns without performing global retrieval. Local overlaps are needed to describe the network. We characterize the structure type of the neural states using a parameter that is related to fluctuations of the local overlaps, with distinction between bump and block phases. Simulation of neural dynamics shows a competition between localized (bump), structured (block) and global retrieval. When the network topology randomness increases, the phase-diagram shows a transition from local to global retrieval. Furthermore, the local phase splits into a bump phase for low activity and a block phase for high activity. A theoretical approach solves the asymptotic limit of the model, and confirms the simulation results which predicts the change of stability from bumps to blocks when the storage ratio increases.

The dynamical activity of a neural network model composed of electrically connected map-based neurons is investigated. After detailing the behavior of the isolated neuron for a wide parameter range, collective network states are depicted using the activity, spatial correlation and time phase distribution as measures. A detailed discussion on the stability of global and partial synchronization states is presented.

The three-state Ising neural network with synchronous updating and variable dilution is discussed starting from the appropriate Hamiltonians. The thermodynamic and retrieval properties are examined using replica mean-field theory. Capacity-temperature phase diagrams are derived for several values of the pattern activity and different gradations of dilution, and the information content is calculated. The results are compared with those for sequential updating. The effect of self-coupling is established. Also the dynamics is studied using the generating function technique for both synchronous and sequential updating. Typical flow diagrams for the overlap order parameter are presented. The differences with the signal-to-noise approach are outlined., Comment: 21 pages Latex, 12 eps figures and 1 ps figure

In the regime of partial synchronization, networks of diffusively coupled Hindmarch–Rose neurons show coherent structures developing in a region of the phase space which is wider than in the correspondent single neuron. Such structures are kept, without important changes, during several bursting periods. In this work, we study the time evolution of these structures and their dynamical stability under damage. This system may model the behavior of ensembles of neurons coupled through a bidirectional gap junction or, in a broader sense, it could also account for the molecular cascades present in the formation of flash and short time memory.

The time evolution of an exactly solvable layered feedforward neural network with three-state neurons and optimizing the mutual information is studied for arbitrary synaptic noise (temperature). Detailed stationary temperature-capacity and capacity-activity phase diagrams are obtained. The model exhibits pattern retrieval, pattern-fluctuation retrieval and spin-glass phases. It is found that there is an improved performance in the form of both a larger critical capacity and information content compared with three-state Ising-type layered network models. Flow diagrams reveal that saddle-point solutions associated with fluctuation overlaps slow down considerably the flow of the network states towards the stable fixed-points., Comment: 17 pages Latex including 6 eps-figures

A recent study on noiseless learning and retrieval in attractor neural networks above saturation, by Griniasty and Gutfreund, is extended to take account for imperfect learning by means of a temperature β −1 = T . Violations of the constraint imposed on the local stabilities are taken into account by various cost functions. The distribution of local stabilities and the fraction of errors during the learning stage are analysed. The retrieval dynamics for the sparsely connected network is studied showing high retrieval overlap in reduced retrieval regions for finite T .

The thermodynamic behavior of an extended Hopfield model for an attractor neural network with units in three Ising-states Si and symmetrically random diluted synaptic connections Jijd, between units i and j, is discussed here in the thermodynamic limit within replica-symmetric mean-field theory. A macroscopic set of random patterns in a two-level hierarchical structure of ancestors and descendents is used to construct the synaptic connections and to determine the magnetic order associated with the recognition of ancestors.

Construction details and calibration techniques used in the implementation of a large lead-glass hodoscope are described. The algorithms used to determine the position and energy of the incident photons as well as the energy and position resolution are discussed.

We report here on recoil implantation of antimony atoms from a 60 nm thickness film deposited on a silicon substrate. The experiments were performed in order to investigate the influence of different implantation parameters on the yield and depth profile of the recoiling antimony atoms. We have used Ar + energies between 40 and 800 keV, doses between 1 × 10 14 and 1 × 10 17 cm −2 , sample temperatures between 20 and 300°C and dose rates between 1 and 20 μA.cm −2 . Another aspect considered here was the process removal of the remaining film. We studied the influence of different conditions of film removal on the recoil yield. The results of the present work indicate that the recoil yield is a function of the energy and total dose of the bombarding ions. Furthermore, no variation with the Ar + dose rate or with the sample temperature was observed. The antimony recoil depth profile depends mainly on the total dose of Ar + and does not depend on its energy, dose rate or target temperature during implantations.