Showing papers in "arXiv: Disordered Systems and Neural Networks in 2003"

Community Structure in Jazz

Abstract: Using a database of jazz recordings we study the collaboration network of jazz musicians. We define the network at two different levels. First we study the collaboration network between individuals, where two musicians are connected if they have played in the same band. Then we consider the collaboration between bands, where two bands are connected if they have a musician in common. The community structure analysis reveals that these constructions capture essential ingredients of the social interactions between jazz musicians. We observe correlations between recording locations, racial segregation and the community structure. A quantitative analysis of the community size distribution reveals a surprising similarity with an e-mail based social network recently studied.

Betweenness Centrality in Large Complex Networks

Abstract: We analyze the betweenness centrality (BC) of nodes in large complex networks. In general, the BC is increasing with connectivity as a power law with an exponent $\eta$. We find that for trees or networks with a small loop density $\eta=2$ while a larger density of loops leads to $\eta<2$. For scale-free networks characterized by an exponent $\gamma$ which describes the connectivity distribution decay, the BC is also distributed according to a power law with a non universal exponent $\delta$. We show that this exponent $\delta$ must satisfy the exact bound $\delta\geq (\gamma+1)/2$. If the scale free network is a tree, then we have the equality $\delta=(\gamma+1)/2$.

Hierarchical Small Worlds in Software Architecture

Abstract: The components of a large software application do not interact in random ways Instead, class diagrams exhibit remarkable topological similarities to other natural and artificial systems The components of a large software application are very well connected because the mean shortest distance between them is very low in spite of having a relatively small number of connections per class In addition, these diagrams are very heterogeneous These measurements are of a general nature and are largely independent of the particular semantics of the application As shown in this paper, and irrespective of the specific features of each system analyzed, the final outcome of software evolution is a small world, hierarchical class diagram with well-defined statistical properties The consequences for software evolution are outlined

Barrier Softening near the onset of Non-Activated Transport in Supercooled Liquids: Implications for Establishing Detailed Connection between Thermodynamic and Kinetic Anomalies in Supercooled Liquids

Abstract: According to the Random First Order Transition (RFOT) theory of glasses, the barriers for activated dynamics in supercooled liquids vanish as the temperature of a viscous liquid approaches the dynamical transition temperature from below. This occurs due to a decrease of the surface tension between local meta-stable molecular arrangements much like at a spinodal. The dynamical transition thus represents a crossover from the low $T$ activated bevavior to a collisional transport regime at high $T$. This barrier softening explains the deviation of the relaxation times, as a function of temperature, from the simple $\log \tau \propto 1/s_c$ dependence at the high viscosity to a mode-mode coupling dominated result at lower viscosity. By calculating the barrier softening effects, the RFOT theory provides a {\em unified} microscopic way to interpret structural relaxation data for many distinct classes of structural glass formers over the measured temperature range. The theory also provides an unambiguous procedure to determine the size of dynamically cooperative regions in the presence of barrier renormalization effects using the experimental temperature dependence of the relaxation times and the configurational entropy data. We use the RFOT theory framework to discuss data for tri-naphthyl benzene, salol, propanol and silica as representative systems.

Thermodynamic Properties of the Dipolar Spin Ice Model

Abstract: We present a detailed theoretical overview of the thermodynamic properties of the dipolar spin ice model, which has been shown to be an excellent quantitative descriptor of the Ising pyrochlore materials Dy_2Ti_2O_7 and Ho_2Ti_2O_7. We show that the dipolar spin ice model can reproduce an effective quasi macroscopically degenerate ground state and spin-ice behavior of these materials when the long-range nature of dipole-dipole interaction is handled carefully using Ewald summation techniques. This degeneracy is, however, ultimately lifted at low temperature. The long-range ordered state is identified via mean field theory and Monte Carlo simulation techniques. Finally, we investigate the behavior of the dipolar spin ice model in an applied magnetic field, and compare our predictions with experimental results. We find that a number of different long-range ordered states are favored by the model depending on field direction.

Dynamics of social networks

Abstract: Complex networks as the World Wide Web, the web of human sexual contacts or criminal networks often do not have an engineered architecture but instead are self-organized by the actions of a large number of individuals. From these local interactions non-trivial global phenomena can emerge as small-world properties or scale-free degree distributions. A simple model for the evolution of acquaintance networks highlights the essential dynamical ingredients necessary to obtain such complex network structures. The model generates highly clustered networks with small average path lengths and scale-free as well as exponential degree distributions. It compares well with experimental data of social networks, as for example coauthorship networks in high energy physics.

The Easiest Hard Problem: Number Partitioning

Abstract: Number partitioning is one of the classical NP-hard problems of combinatorial optimization. It has applications in areas like public key encryption and task scheduling. The random version of number partitioning has an "easy-hard" phase transition similar to the phase transitions observed in other combinatorial problems like $k$-SAT. In contrast to most other problems, number partitioning is simple enough to obtain detailled and rigorous results on the "hard" and "easy" phase and the transition that separates them. We review the known results on random integer partitioning, give a very simple derivation of the phase transition and discuss the algorithmic implications of both phases.

Network dynamics of ongoing social relationships

Abstract: Many recent large-scale studies of interaction networks have focused on networks of accumulated contacts. In this paper we explore social networks of ongoing relationships with an emphasis on dynamical aspects. We find a distribution of response times (times between consecutive contacts of different direction between two actors) that has a power-law shape over a large range. We also argue that the distribution of relationship duration (the time between the first and last contacts between actors) is exponentially decaying. Methods to reanalyze the data to compensate for the finite sampling time are proposed. We find that the degree distribution for networks of ongoing contacts fits better to a power-law than the degree distribution of the network of accumulated contacts do. We see that the clustering and assortative mixing coefficients are of the same order for networks of ongoing and accumulated contacts, and that the structural fluctuations of the former are rather large.

Low-temperature dynamics of weakly localized Frenkel excitons in disordered linar chains

Abstract: We calculate the temperature dependence of the fluorescence Stokes shift and the fluorescence decay time in linear Frenkel exciton systems resulting from the thermal redistribution of exciton population over the band states. The following factors, relevant to common experimental conditions, are accounted for in our kinetic model: (weak) localization of the exciton states by static disorder, coupling of the localized excitons to vibrations in the host medium, a possible non-equilibrium of the subsystem of localized Frenkel excitons on the time scale of the emission process, and different excitation conditions (resonant or non resonant). A Pauli master equation, with microscopically calculated transition rates, is used to describe the redistribution of the exciton population over the manifold of localized exciton states. We find a counterintuitive non-monotonic temperature dependence of the Stokes shift. In addition, we show that depending on experimental conditions, the observed fluorescence decay time may be determined by vibration-induced intra-band relaxation, rather than radiative relaxation to the ground state. The model considered has relevance to a wide variety of materials, such as linear molecular aggregates, conjugated polymers, and polysilanes.

Recent Developments of World-Line Monte Carlo Methods

Abstract: World-line quantum Monte Carlo methods are reviewed with an emphasis on breakthroughs made in recent years. In particular, three algorithms -- the loop algorithm, the worm algorithm, and the directed-loop algorithm -- for updating world-line configurations are presented in a unified perspective. Detailed descriptions of the algorithms in specific cases are also given.

A computational approach to regulatory element discovery in eukaryotes

Abstract: Gene regulation in Eukaryotes is mainly effected through transcription factors binding to rather short recognition motifs generally located upstream of the coding region. We present a novel computational method to identify regulatory elements in the upstream region of Eukaryotic genes. The genes are grouped in sets sharing an overrepresented short motif in their upstream sequence. For each set, the average expression level from a microarray experiment is determined: if this level is significantly higher or lower than the average taken over the whole genome, then the overrepresented motif shared by the genes in the set is likely to play a role in their regulation. We illustrate the method by applying it to the genome of {\it S. cerevisiae}, for which many datasets of microarray experiments are publicly available. Several known binding motifs are correctly recognized by our algorithm, and a new candidate is suggested for experimental verification.

Recent Progress in Spin Glasses

Abstract: We review recent findings on spin glass models. Both the equilibrium properties and the dynamic properties are covered. We focus on progress in theoretical, in particular numerical, studies, while its relationship to real magnetic materials is also mentioned.

Low-energy excess of vibrational states in v-SiO2: the role of transverse dynamics

Abstract: The study of the effects of the density variations on the vibrational dynamics in vitreous silica is presented. A detailed analysis of the dynamical structure factor, as well as of the current spectra, allows the identification of a flattened transverse branch which is highly sensitive to the density variations. The experimental variations on the intensity and position of the Boson Peak (BP) in v-SiO2 as a function of density are reproduced and interpreted as being due to the shift and disappearance of the latter band. The BP itself is found to correspond to the lower energy tail of the excess states due to the piling up of vibrational modes at energies corresponding to the flattening of the transverse branch.

Using Recurrent Neural Networks To Forecasting of Forex

Abstract: This paper reports empirical evidence that a neural networks model is applicable to the statistically reliable prediction of foreign exchange rates. Time series data and technical indicators such as moving average, are fed to neural nets to capture the underlying "rules" of the movement in currency exchange rates. The trained recurrent neural networks forecast the exchange rates between American Dollar and four other major currencies, Japanese Yen, Swiss Frank, British Pound and EURO. Various statistical estimates of forecast quality have been carried out. Obtained results show, that neural networks are able to give forecast with coefficient of multiple determination not worse then 0.65. Linear and nonlinear statistical data preprocessing, such as Kolmogorov-Smirnov test and Hurst exponents for each currency were calculated and analyzed.

Chiral susceptibility in canonical spin glass and reentrant alloys from Hall effect measurements

Abstract: The extraordinary Hall effect coefficients $R_s(T)$ of the canonical spin glass alloys ${\bf Au}Fe 8 at%$ and ${\bf Au}Mn 8 at%$ and of an archetype reentrant system ${\bf Au}Fe 18 at%$ were measured as functions of temperature. The data show a critical cusp-like term superimposed on a smooth background for the spin glasses, and change of sign with temperature for the reentrant. The results can be interpreted consistently by invoking a chiral Hall effect contribution as proposed by Kawamura.

Influence of quenched dilution on the quasi-long-range ordered phase of the 2d XY model

Abstract: The influence of non magnetic impurities in the 2d XY model is investigated through Monte Carlo (MC) simulations. The general picture of the transition is fully understood from the Harris criterion which predicts that the universality class is unchanged, and the Berezinskii-Kosterlitz-Thouless description of the topological transition remains valid. We nevertheless address here the question about the influence of dilution on the quasi-long-range order at low temperatures. In particular, we study the asymptotic of the pair correlation function and report the MC estimates for the critical exponent $\eta$ at different dilutions. In the weak dilution region, our MC calculations are further supported by simple spin-wave-like calculations.

Off-equilibrium dynamics of the two-dimensional Coulomb glass

Abstract: The dynamics of the 2D Coulomb glass model is investigated by kinetic Monte Carlo simulation. An exponential divergence of the relaxation time signals a zero-temperature freezing transition. At low temperatures the dynamics of the system is glassy. The local charge correlations and the response to perturbations of the local potential show aging. The dynamics of formation of the Coulomb gap is slow and the density of states at the Fermi level decays in time as a power law. The relevance of these findings for recent transport experiments in Anderson-insulating films is pointed out.

Ground-state energy fluctuations in the Sherrington-Kirkpatrick model

Abstract: The probability distribution function (PDF) of the ground-state energy in the Sherrington-Kirkpatrick spin-glass model is numerically determined by collecting a large statistical sample of ground states, computed using a genetic algorithm. It is shown that the standard deviation of the ground-state energy per spin scales with the number of spins, N, as N^{-\rho} with \rho \simeq 0.765, but the value \rho=3/4 is also compatible with the data, while the previously proposed value \rho=5/6 is ruled out. The PDF satisfies finite-size scaling with a non-Gaussian asymptotic PDF, which can be fitted remarkably well by the Gumbel distribution for the m-th smallest element in a set of random variables, with m \simeq 6.

Low-Temperature Excitations of Dilute Lattice Spin Glasses

Abstract: A new approach to exploring low-temperature excitations in finite-dimensional lattice spin glasses is proposed. By focusing on bond-diluted lattices just above the percolation threshold, large system sizes $L$ can be obtained which lead to enhanced scaling regimes and more accurate exponents. Furthermore, this method in principle remains practical for any dimension, yielding exponents that so far have been elusive. This approach is demonstrated by determining the stiffness exponent for dimensions $d=3$, $d=6$ (the upper critical dimension), and $d=7$. Key is the application of an exact reduction algorithm, which eliminates a large fraction of spins, so that the reduced lattices never exceed $\sim10^3$ variables for sizes as large as L=30 in $d=3$, L=9 in $d=6$, or L=8 in $d=7$. Finite size scaling analysis gives $y_3=0.24(1)$ for $d=3$, significantly improving on previous work. The results for $d=6$ and $d=7$, $y_6=1.1(1)$ and $y_7=1.24(5)$, are entirely new and are compared with mean-field predictions made for d>=6.

Virial expansion for almost diagonal random matrices

Abstract: Energy level statistics of Hermitian random matrices $\hat H$ with Gaussian independent random entries $H_{i\geq j}$ is studied for a generic ensemble of almost diagonal random matrices with $ \sim 1$ and $ = b {\cal F}(|i-j|) \ll 1$. We perform a regular expansion of the spectral form-factor $K(\tau) = 1 + b K_{1}(\tau) + b^{2} K_{2}(\tau) + ... $ in powers of $b \ll 1$ with the coefficients $K_{m}(\tau)$ that take into account interaction of (m+1) energy levels. To calculate $K_{m}(\tau)$, we develop a diagrammatic technique which is based on the Trotter formula and on the combinatorial problem of graph edges coloring with (m+1) colors. Expressions for $K_{1}(\tau)$ and $K_{2}(\tau)$ in terms of infinite series are found for a generic function ${\cal F}(|i-j|)$ in the Gaussian Orthogonal Ensemble (GOE), the Gaussian Unitary Ensemble (GUE) and in the crossover between them (the almost unitary Gaussian ensemble). The Rosenzweig-Porter and power-law banded matrix ensembles are considered as examples.

The "Cameo Principle" and the Origin of Scale-Free Graphs in Social Networks

Abstract: We formulate a simple edge generation rule based on an inverse like mass action principle for random graphs over a structured vertex set. We show that under very weak assumptions on the structure generating distribution we obtain a scale free distribution for the degree. We furthermore introduce and study a "my friends are your friends" local search principle which makes the clustering coefficient large.

Maximum matching on random graphs

Abstract: The maximum matching problem on random graphs is studied analytically by the cavity method of statistical physics. When the average vertex degree \mth{c} is larger than \mth{2.7183}, groups of max-matching patterns which differ greatly from each other {\em gradually} emerge. An analytical expression for the max-matching size is also obtained, which agrees well with computer simulations. Discussion is made on this {\em continuous} glassy phase transition and the absence of such a glassy phase in the related minimum vertex covering problem.

Field-Shift Aging Protocol on the 3D Ising Spin-Glass Model: Dynamical Crossover between the Spin-Glass and Paramagnetic States

Abstract: Spin-glass (SG) states of the 3-dimensional Ising Edwards-Anderson model under a static magnetic field $h$ are examined by means of the standard Monte Carlo simulation on the field-shift aging protocol at temperature $T$. For each process with $(T; \tw, h)$, $\tw$ being the waiting time before the field is switched on, we extract the dynamical crossover time, $\tcr(T; \tw, h)$. We have found a nice scaling relation between the two characteristic length scales which are properly determined from $\tcr$ and $\tw$ and then are normalized by the static field crossover length introduced in the SG droplet theory. This scaling behavior implies the instability of the SG phase in the equilibrium limit even under an infinitesimal $h$. In comparison with this numerical result the field effect on real spin glasses is also discussed.

Seven clusters in genomic triplet distributions

Abstract: In several recent papers new gene-detection algorithms were proposed for detecting protein-coding regions without requiring learning dataset of already known genes. The fact that unsupervised gene-detection is possible closely connected to existence of a cluster structure in oligomer frequency distributions. In this paper we study cluster structure of several genomes in the space of their triplet frequencies, using pure data exploration strategy. Several complete genomic sequences were analyzed, using visualization of tables of triplet frequencies in a sliding window. The distribution of 64-dimensional vectors of triplet frequencies displays a well-detectable cluster structure. The structure was found to consist of seven clusters, corresponding to protein-coding information in three possible phases in one of the two complementary strands and in the non-coding regions with high accuracy (higher than 90% on the nucleotide level). Visualizing and understanding the structure allows to analyze effectively performance of different gene-prediction tools. Since the method does not require extraction of ORFs, it can be applied even for unassembled genomes. The information content of the triplet distributions and the validity of the mean-field models are analysed.

Combining Hebbian and reinforcement learning in a minibrain model

Abstract: A toy model of a neural network in which both Hebbian learning and reinforcement learning occur is studied. The problem of `path interference', which makes that the neural net quickly forgets previously learned input-output relations is tackled by adding a Hebbian term (proportional to the learning rate $\eta$) to the reinforcement term (proportional to $\rho$) in the learning rule. It is shown that the number of learning steps is reduced considerably if $1/4 < \eta/\rho < 1/2$, i.e., if the Hebbian term is neither too small nor too large compared to the reinforcement term.

Potential Energy Landscape in Lennard-Jones binary mixture model

Abstract: The potential energy landscape in the Kob-Andersen Lennard-Jones binary mixture model has been studied carefully from liquid down to the supercooled regime, from T =2 down to T =0.46. One thousand of independent configurations along the time evolution have been examined at each investigated temperature. From the starting configuration we searched the nearest saddle (or quasi-saddle) and minimum of the potential energy. The vibrational densities of states for the starting and the two derived configurations have been evaluated. Besides the number of negative eigenvalues of saddle, other quantities show some signature of the approaching of the dynamical arrest temperature.

A backtracking survey propagation algorithm for K-satisfiability

Abstract: In this paper we present a backtracking version of the survey propagation algorithm We show that the introduction of the simplest form of backtracking greatly improves the ability of the original survey propagation algorithm in solving difficult random problems near the sat-unsat transition

Spin Glasses: Still Complex After All These Years?

Abstract: Spin glasses are magnetic systems exhibiting both quenched disorder and frustration, and have often been cited as examples of `complex systems.' In this talk I review some of the basic notions of spin glass physics, and discuss how some of our recent progress in understanding their properties might lead to new viewpoints of how they manifest `complexity'.

Bicoloring Random Hypergraphs

Abstract: We study the problem of bicoloring random hypergraphs, both numerically and analytically. We apply the zero-temperature cavity method to find analytical results for the phase transitions (dynamic and static) in the 1RSB approximation. These points appear to be in agreement with the results of the numerical algorithm. In the second part, we implement and test the Survey Propagation algorithm for specific bicoloring instances in the so called HARD-SAT phase.

Cut-offs and finite size effects in scale-free networks

Abstract: We analyze the degree distribution's cut-off in finite size scale-free networks. We show that the cut-off behavior with the number of vertices $N$ is ruled by the topological constraints induced by the connectivity structure of the network. Even in the simple case of uncorrelated networks, we obtain an expression of the structural cut-off that is smaller that the natural cut-off obtained by means of extremal theory arguments. The obtained results are explicitly applied in the case of the configuration model to recover the size scaling of tadpoles and multiple edges.