Bayesian model selection for time series using Markov chain Monte Carlo

We present a stochastic simulation technique for subset selection in time series models, based on the use of indicator variables with the Gibbs sampler within a hierarchical Bayesian framework. As an example, the method is applied to the selection of subset linear AR models, in which only significant lags are included. Joint sampling of the indicators and parameters is found to speed convergence. We discuss the possibility of model mixing where the model is not well determined by the data, and the extension of the approach to include non-linear model terms.

Sequential Monte Carlo computation of the score and observed information matrix in state-space models with application to parameter estimation

Sequential Monte Carlo (SMC) methods are popular computational tools for Bayesian inference in non-linear non-Gaussian state-space models. For this class of models, we propose SMC algorithms to compute the score vector and observed information matrix recursively in time. We propose two different SMC implementations, one with computational complexity $\mathcal{O}(N)$ and the other with complexity $\mathcal{O}(N^{2})$ where $N$ is the number of importance sampling draws. Although cheaper, the performance of the $\mathcal{O}(N)$ method degrades quickly in time as it inherently relies on the SMC approximation of a sequence of probability distributions whose dimension is increasing linearly with time. In particular, even under strong \textit{mixing} assumptions, the variance of the estimates computed with the $\mathcal{O}(N)$ method increases at least quadratically in time. The $\mathcal{O}(N^{2})$ is a non-standard SMC implementation that does not suffer from this rapid degrade. We then show how both methods can be used to perform batch and recursive parameter estimation.

An approximate likelihood method for estimating static parameters in multi-target tracking models

This paper considers a class of dynamic Spatial Point Processes (PP) that evolves over time in a Markovian fashion. This Markov in time PP is hidden and observed indirectly through another PP via thinning, displacement and noise. This statistical model is important for Multi object Tracking applications and we present an approximate likelihood based method for estimating the model parameters. The work is supported by an extensive numerical study.

The mechanism of pattern selection in coupled map lattices

The pattern selection of one-dimensional coupled map lattices is studied in this paper. It is shown by spatiotemporal variable separation that there exists a threshold wavelength in pattern selection which possesses wave-like structures in space and periodic chaotic motion in time.

Model of sources: a proposal for the hierarchy, merging strategy and representation of the information sources in virtual models of historical buildings

Contributed to: Fusion of Cultures: XXXVIII Annual Conference on Computer Applications and Quantitative Methods in Archaeology – CAA2010 (Granada, Spain, Apr 6-9, 2010)

Para una relectura del compromiso en los años sesenta: paradoja, escritura transparente, retracción de Antonio Gamoneda

[ES] En este artículo se trata de analizar el planteamiento del compromiso poético a la altura del comienzo de la década de los años sesenta, a partir del análisis de esa modalidad de escritura en Blues castellano de Antonio Gamoneda, libro cuyos poemas datan de 1961–1966, aunque permaneció inédito hasta 1982. Leído en el contexto de su producción, Blues castellano muestra una solidaridad absoluta con algunos de los modelos poéticos comprometidos de la época (Nazim Hikmet, Bertolt Brecht, los espirituales negros, etc.), así como una vinculación hacia un tipo de compromiso poético que va a hacer recaer su dimensión crítica en el lenguaje poético, no sólo en el referente externo. En Blues castellano se consolidan ya algunos de los conceptos centrales de la poesía gamonediana posterior: retracción, búsqueda de una escritura transparente, desaparición del sujeto en la escritura, etc.

Expected Fair Allocation in Farsighted Network Formation

I consider cooperation situations where players have network relations. Networks evolve according to a stationary transition probability matrix and at each moment in time players receive payoffs from a stationary allocation rule. Players discount the future by a common factor. The pair formed by an allocation rule and a transition probability matrix is called expected fair if for every link in the network both participants gain, marginally, and in discounted, expected terms, the same from it; and it is called a pairwise network formation procedure if the probability that a link is created (or eliminated) is positive if the discounted, expected gains to its two participants are positive too. The main result is the existence, for the discount factor small enough, of an expected fair and pairwise network formation procedure where the allocation rule is component balanced, meaning it distributes the total value of any maximal connected subnetwork among its participants. This existence result holds for all discount factors when the pairwise network formation procedure is restricted. I finally provide some comparison with previous models of farsighted network formation.

Forward-looking Pairwise Stability in Networks with Externalities

We consider cooperation situations where players have network relations. Networks evolve according to a stationary transition probability matrix and at each moment in time players receive payoffs from a stationary allocation rule. Players discount the future by a common factor. The pair formed by an allocation rule and a transition probability matrix is called a forward-looking network formation scheme if, first, the probability that a link is created is positive if the discounted, expected gains to its two participants are positive, and if, second, the probability that a link is eliminated is positive if the discounted, expected gains to at least one of its two participants are positive. The main result is the existence, for all discount factors and all value functions, of a forward-looking network formation scheme. Furthermore, we can always nd a forward-looking network formation scheme such that (i) the allocation rule is component balanced and (ii) the transition probabilities increase in the di erence in payo s for the corresponding players responsible for the transition. We use this dynamic solution concept to explore the tension between e ciency and stability.

Distribution, relative abundance and developmental morphology of paralarval cephalopods in the Western North Atlantic Ocean

Paralarval and juvenile cephalopods collected in plankton samples on 21 western North Atlantic cruises were identified and enumerated. The 3731 specimens were assigned to 44 generic and specific taxa. This paper describes their spatial and temporal distributions and their developmental morphology. The smallest paralarvae recognized for a number of species are identified and illustrated. The two most abundant and most frequently collected taxa were identifiable to species based on known systematic characters of young, as well as on distribution of the adults. These were the neritic squids Loligo pealeii and Illex illecebrosus collected north of Cape Hatteras, both valuable fishery resources. Other abundant taxa included two morphotypes of ommastrephids, at least five species of enoploteuthids, two species of onychoteuthids, and unidentified octopods. Most taxa were distributed widely both in time and in space, although some seasonal and mesoscale-spatial patterns were indicated. The taxa that appeared to have distinct seasonal distribution included most of the neritic species and, surprisingly, the young of the bathypelagic cranchiids. In eight seasonal cruises over the continental shelf of the middle U.S. Atlantic states, neritic taxa demonstrated approximately the same seasonal patterns during two consecutive years. Interannual differences in the oceanic taxa collected on the shelf were extreme. The highest abundance and diversity of planktonic cephalopods in the oceanic samples were consistently found in the vicinity of the Gulf Stream. Only eight of the oceanic taxa appeared to have limited areal distributions, compared with twelve taxa that were found throughout the western North Atlantic regions sampled in this study. Many taxa, however, were not collected frequently enough to describe seasonal or spatial patterns. Comparisons with published accounts of other cephalopod surveys indicate both strengths and weaknesses in various sampling techniques for capturing the young of oceanic cephalopods. Enoploteuthids were abundant both in our study and in other studies using midwater trawls in several areas of the North Atlantic. Thus, this family probably is adequately sampled over its developmental range. In contrast, octopoteuthids and chtenopterygiids are rare in collections made by small to medium-sized midwater trawls but are comparatively common in plankton samples. For families that are relatively common in plankton samples, paralarval abundance, derived similarly to the familiar ichthyoplankton surveys of fisheries science, may be the most reliable method of gathering data on distribution and abundance. (PDF file contains 58 pages.)

Community structure in real-world networks from a non-parametrical synchronization-based dynamical approach

[EN]This work analyzes the problem of community structure in real-world networks based on the synchronization of nonidentical coupled chaotic Rössler oscillators each one characterized by a defined natural frequency, and coupled according to a predefined network topology. The interaction scheme contemplates an uniformly increasing coupling force to simulate a society in which the association between the agents grows in time. To enhance the stability of the correlated states that could emerge from the synchronization process, we propose a parameterless mechanism that adapts the characteristic frequencies of coupled oscillators according to a dynamic connectivity matrix deduced from correlated data. We show that the characteristic frequency vector that results from the adaptation mechanism reveals the underlying community structure present in the network.

Contributions to unsupervised and supervised learning with applications in digital image processing

311 p. : il.

Netzmaterial und Selektivität in der Ostseedorschfischerei

Since some time fishing gear scientists express their concern over an observed tendency of the commercial fishery to proceed from codend netting yarns of 3 to 4 mm to higher values or to switch to the use of double instead of single yarn. A recent large EU-financed project collected statistical evidence on the detrimental effect of such behaviour on the selectivity of the codends. In this context data on cod are very scarce. German-Polish experiments in the Baltic from 1999 to 2001 aimed at filling this gap. The investigations prove a clear evidence of a negativ ecorrelation between netting yarn diameter and selectivity factor and/or L50. In addition they demonstrate a clear negative effect on selectivity when switching from single to double yarn The effects are of an order of magnitude that counteracting effects as catch size are masked and support the decision of the IBSFC to define maximum yarn diameters both for single and double yarn netting. A measuring instrument for the enforcement of these new regulations was introduced right in time.

Bestimmung von künstlichen Radionukliden in Konsumfisch

The identification of artificial radionuclides in fish involves some diffculties, because the quantities of these nuclides are very low (10-16 to 10-10 g/kg). The procedures have to be done very carefully. The sample preparation, the radiochemical analyses and the final preparation of the samples for the detection of the radioactivity of strontium-90, plutonium-238, -239, -240 and americium-241 are briefly described. The levels of artificial radioactivity in some species of fish from the North Sea are shown. The additional exposure to radiation by artificial radionuc1ides by ingestion of fish amounts only to about 0,02 % of the mean exposure to natural radiation. Nevertheless further monitoring of radioactivity should be continued in order to ensure that changes can be detected in time.

Erster Einsatz von Trenngittern in der kommerziellen Garnelenfischerei

Comparative fishing trials were conducted in the river Elbe estuary using 9 m commercial brown shrimp beam trawls. To avoid the bycatch of fish a metal sorting grid of the Nordmöre type was used. The elliptical grid was constructed of 6 mm stainless steel bar with a spacing of 20 mm between the bars and housed in a cylindrical frame of 800 mm diameter. It was installed in the extension piece just in front of the codend. The inclination of the grid was 45 degrees. A fish outlet was provided in the upper panel of the trawl at the upper edge of the grid. A series of 8 tows of 15 min duration at a towing speed of 3 kns was done. For evaluation the catch of the main codend was compared to the portion of the catch escaped through the grid. The presence of the grid caused a 97.4 % reduction of the catch of lump sucker, a 90.6 % reduction of the catch of sea scorpion, a 79.3 % reduction of the catch of cod, a 58.8 % reduction of the catch of armed bullhead, a 39.6 % reduction of the catch of dab, a 34.7 % reduction of the catch of flounder, a 32.3 % reduction of the catch of smelt, a 19.8 % reduction of the catch of plaice and a 14.5 % reduction of the catch of brown shrimp.

Nonlinear oscillations in discrete and continuous systems

The first thesis topic is a perturbation method for resonantly coupled nonlinear oscillators. By successive near-identity transformations of the original equations, one obtains new equations with simple structure that describe the long time evolution of the motion. This technique is related to two-timing in that secular terms are suppressed in the transformation equations. The method has some important advantages. Appropriate time scalings are generated naturally by the method, and don't need to be guessed as in two-timing. Furthermore, by continuing the procedure to higher order, one extends (formally) the time scale of valid approximation. Examples illustrate these claims. Using this method, we investigate resonance in conservative, non-conservative and time dependent problems. Each example is chosen to highlight a certain aspect of the method.

The second thesis topic concerns the coupling of nonlinear chemical oscillators. The first problem is the propagation of chemical waves of an oscillating reaction in a diffusive medium. Using two-timing, we derive a nonlinear equation that determines how spatial variations in the phase of the oscillations evolves in time. This result is the key to understanding the propagation of chemical waves. In particular, we use it to account for certain experimental observations on the Belusov-Zhabotinskii reaction.

Next, we analyse the interaction between a pair of coupled chemical oscillators. This time, we derive an equation for the phase shift, which measures how much the oscillators are out of phase. This result is the key to understanding M. Marek's and I. Stuchl's results on coupled reactor systems. In particular, our model accounts for synchronization and its bifurcation into rhythm splitting.

Finally, we analyse large systems of coupled chemical oscillators. Using a continuum approximation, we demonstrate mechanisms that cause auto-synchronization in such systems.