CADS*: Computer-Aided Development of self-* systems

We present the prototype tool CADS* for the computer-aided development of an important class of self-* systems, namely systems whose components can be modelled as Markov chains. Given a Markov chain representation of the IT components to be included into a self-* system, CADS* automates or aids (a) the development of the artifacts necessary to build the self-* system; and (b) their integration into a fully-operational self-* solution. This is achieved through a combination of formal software development techniques including model transformation, model-driven code generation and dynamic software reconfiguration.

An accurate and scalable analytical model for IEEE 802.15.4 slotted CSMA/CA networks

In this paper a Markov chain based analytical model is proposed to evaluate the slotted CSMA/CA algorithm specified in the MAC layer of IEEE 802.15.4 standard. The analytical model consists of two two-dimensional Markov chains, used to model the state transition of an 802.15.4 device, during the periods of a transmission and between two consecutive frame transmissions, respectively. By introducing the two Markov chains a small number of Markov states are required and the scalability of the analytical model is improved. The analytical model is used to investigate the impact of the CSMA/CA parameters, the number of contending devices, and the data frame size on the network performance in terms of throughput and energy efficiency. It is shown by simulations that the proposed analytical model can accurately predict the performance of slotted CSMA/CA algorithm for uplink, downlink and bi-direction traffic, with both acknowledgement and non-acknowledgement modes.

One Class of Stochastic Local Search Algorithms

Accelerated probabilistic modeling algorithms, presenting stochastic local search (SLS) technique, are considered. General algorithm scheme and specific combinatorial optimization method, using “golden section” rule (GS-method), are given. Convergence rates using Markov chains are received. An overview of current combinatorial optimization techniques is presented.

Quasi-Monte Carlo Methods for some Linear Algebra Problems. Convergence and Complexity

We present quasi-Monte Carlo analogs of Monte Carlo methods for some linear algebra problems: solving systems of linear equations, computing extreme eigenvalues, and matrix inversion. Reformulating the problems as solving integral equations with a special kernels and domains permits us to analyze the quasi-Monte Carlo methods with bounds from numerical integration. Standard Monte Carlo methods for integration provide a convergence rate of O(N^(−1/2)) using N samples. Quasi-Monte Carlo methods use quasirandom sequences with the resulting convergence rate for numerical integration as good as O((logN)^k)N^(−1)). We have shown theoretically and through numerical tests that the use of quasirandom sequences improves both the magnitude of the error and the convergence rate of the considered Monte Carlo methods. We also analyze the complexity of considered quasi-Monte Carlo algorithms and compare them to the complexity of the analogous Monte Carlo and deterministic algorithms.

Maximal Number of Successors in a Nginar(1) Process

2000 Mathematics Subject Classification: 60J80, 60J20, 60J10, 60G10, 60G70, 60F99.

A Note on the Extinction Problem for Controlled Multitype Branching Processes

2000 Mathematics Subject Classification: 60J80, 60J10.

Using observation ageing to improve Markovian model learning in QoS engineering

Markovian models are widely used to analyse quality-of-service properties of both system designs and deployed systems. Thanks to the emergence of probabilistic model checkers, this analysis can be performed with high accuracy. However, its usefulness is heavily dependent on how well the model captures the actual behaviour of the analysed system. Our work addresses this problem for a class of Markovian models termed discrete-time Markov chains (DTMCs). We propose a new Bayesian technique for learning the state transition probabilities of DTMCs based on observations of the modelled system. Unlike existing approaches, our technique weighs observations based on their age, to account for the fact that older observations are less relevant than more recent ones. A case study from the area of bioinformatics workflows demonstrates the effectiveness of the technique in scenarios where the model parameters change over time.

Secure information flow via stripping and fast simulation

Type systems for secure information flow aim to prevent a program from leaking information from H (high) to L (low) variables. Traditionally, bisimulation has been the prevalent technique for proving the soundness of such systems. This work introduces a new proof technique based on stripping and fast simulation, and shows that it can be applied in a number of cases where bisimulation fails. We present a progressive development of this technique over a representative sample of languages including a simple imperative language (core theory), a multiprocessing nondeterministic language, a probabilistic language, and a language with cryptographic primitives. In the core theory we illustrate the key concepts of this technique in a basic setting. A fast low simulation in the context of transition systems is a binary relation where simulating states can match the moves of simulated states while maintaining the equivalence of low variables; stripping is a function that removes high commands from programs. We show that we can prove secure information flow by arguing that the stripping relation is a fast low simulation. We then extend the core theory to an abstract distributed language under a nondeterministic scheduler. Next, we extend to a probabilistic language with a random assignment command; we generalize fast simulation to the setting of discrete time Markov Chains, and prove approximate probabilistic noninterference. Finally, we introduce cryptographic primitives into the probabilistic language and prove computational noninterference, provided that the underling encryption scheme is secure.

Behavioural profiles of captive capuchin monkeys (Sapajus spp.): analyses at group and individual levels

The use of behavioural indicators of suffering and welfare in captive animals has produced ambiguous results. In comparisons between groups, those in worse condition tend to exhibit increased overall rate of Behaviours Potentially Indicative of Stress (BPIS), but when comparing within groups, individuals differ in their stress coping strategies. This dissertation presents analyses to unravel the Behavioural Profile of a sample of 26 captive capuchin monkeys, of three different species (Sapajus libidinosus, S. flavius and S. xanthosternos), kept in different enclosure types. In total, 147,17 hours of data were collected. We explored four type of analysis: Activity Budgets, Diversity indexes, Markov chains and Sequence analyses, and Social Network Analyses, resulting in nine indexes of behavioural occurrence and organization. In chapter One we explore group differences. Results support predictions of minor sex and species differences and major differences in behavioural profile due to enclosure type: i. individuals in less enriched enclosures exhibited a more diverse BPIS repertoire and a decreased probability of a sequence with six Genus Normative Behaviour; ii. number of most probable behavioural transitions including at least one BPIS was higher in less enriched enclosures; iii. proeminence indexes indicate that BPIS function as dead ends of behavioural sequences, and proeminence of three BPIS (pacing, self-direct, active I) were higher in less enriched enclosures. Overall, these data are not supportive of BPIS as a repetitive pattern, with a mantra-like calming effect. Rather, the picture that emerges is more supportive of BPIS as activities that disrupt organization of behaviours, introducing “noise” that compromises optimal activity budget. In chapter Two we explored individual differences in stress coping strategies. We classified individuals along six axes of exploratory behaviour. These were only weakly correlated indicating low correlation among behavioural indicators of syndromes. Nevertheless, the results are suggestive of two broad stress coping strategies, similar to the bold/proactive and shy/reactive pattern: more exploratory capuchin monkeys exhibited increased values of proeminence in Pacing, aberrant sexual display and Active 1 BPIS, while less active animals exhibited increased probability in significant sequences involving at least one BPIS, and increased prominence in own stereotypy. Capuchin monkeys are known for their cognitive capacities and behavioural flexibility, therefore, the search for a consistent set of behavioural indictors of welfare and individual differences requires further studies and larger data sets. With this work we aim contributing to design scientifically grounded and statistically correct protocols for collection of behavioural data that permits comparability of results and meta-analyses, from whatever theoretical perspective interpretation it may receive.

Bayesian Inference in Large-scale Problems

Many modern applications fall into the category of "large-scale" statistical problems, in which both the number of observations n and the number of features or parameters p may be large. Many existing methods focus on point estimation, despite the continued relevance of uncertainty quantification in the sciences, where the number of parameters to estimate often exceeds the sample size, despite huge increases in the value of n typically seen in many fields. Thus, the tendency in some areas of industry to dispense with traditional statistical analysis on the basis that "n=all" is of little relevance outside of certain narrow applications. The main result of the Big Data revolution in most fields has instead been to make computation much harder without reducing the importance of uncertainty quantification. Bayesian methods excel at uncertainty quantification, but often scale poorly relative to alternatives. This conflict between the statistical advantages of Bayesian procedures and their substantial computational disadvantages is perhaps the greatest challenge facing modern Bayesian statistics, and is the primary motivation for the work presented here.

Two general strategies for scaling Bayesian inference are considered. The first is the development of methods that lend themselves to faster computation, and the second is design and characterization of computational algorithms that scale better in n or p. In the first instance, the focus is on joint inference outside of the standard problem of multivariate continuous data that has been a major focus of previous theoretical work in this area. In the second area, we pursue strategies for improving the speed of Markov chain Monte Carlo algorithms, and characterizing their performance in large-scale settings. Throughout, the focus is on rigorous theoretical evaluation combined with empirical demonstrations of performance and concordance with the theory.

One topic we consider is modeling the joint distribution of multivariate categorical data, often summarized in a contingency table. Contingency table analysis routinely relies on log-linear models, with latent structure analysis providing a common alternative. Latent structure models lead to a reduced rank tensor factorization of the probability mass function for multivariate categorical data, while log-linear models achieve dimensionality reduction through sparsity. Little is known about the relationship between these notions of dimensionality reduction in the two paradigms. In Chapter 2, we derive several results relating the support of a log-linear model to nonnegative ranks of the associated probability tensor. Motivated by these findings, we propose a new collapsed Tucker class of tensor decompositions, which bridge existing PARAFAC and Tucker decompositions, providing a more flexible framework for parsimoniously characterizing multivariate categorical data. Taking a Bayesian approach to inference, we illustrate empirical advantages of the new decompositions.

Latent class models for the joint distribution of multivariate categorical, such as the PARAFAC decomposition, data play an important role in the analysis of population structure. In this context, the number of latent classes is interpreted as the number of genetically distinct subpopulations of an organism, an important factor in the analysis of evolutionary processes and conservation status. Existing methods focus on point estimates of the number of subpopulations, and lack robust uncertainty quantification. Moreover, whether the number of latent classes in these models is even an identified parameter is an open question. In Chapter 3, we show that when the model is properly specified, the correct number of subpopulations can be recovered almost surely. We then propose an alternative method for estimating the number of latent subpopulations that provides good quantification of uncertainty, and provide a simple procedure for verifying that the proposed method is consistent for the number of subpopulations. The performance of the model in estimating the number of subpopulations and other common population structure inference problems is assessed in simulations and a real data application.

In contingency table analysis, sparse data is frequently encountered for even modest numbers of variables, resulting in non-existence of maximum likelihood estimates. A common solution is to obtain regularized estimates of the parameters of a log-linear model. Bayesian methods provide a coherent approach to regularization, but are often computationally intensive. Conjugate priors ease computational demands, but the conjugate Diaconis--Ylvisaker priors for the parameters of log-linear models do not give rise to closed form credible regions, complicating posterior inference. In Chapter 4 we derive the optimal Gaussian approximation to the posterior for log-linear models with Diaconis--Ylvisaker priors, and provide convergence rate and finite-sample bounds for the Kullback-Leibler divergence between the exact posterior and the optimal Gaussian approximation. We demonstrate empirically in simulations and a real data application that the approximation is highly accurate, even in relatively small samples. The proposed approximation provides a computationally scalable and principled approach to regularized estimation and approximate Bayesian inference for log-linear models.

Another challenging and somewhat non-standard joint modeling problem is inference on tail dependence in stochastic processes. In applications where extreme dependence is of interest, data are almost always time-indexed. Existing methods for inference and modeling in this setting often cluster extreme events or choose window sizes with the goal of preserving temporal information. In Chapter 5, we propose an alternative paradigm for inference on tail dependence in stochastic processes with arbitrary temporal dependence structure in the extremes, based on the idea that the information on strength of tail dependence and the temporal structure in this dependence are both encoded in waiting times between exceedances of high thresholds. We construct a class of time-indexed stochastic processes with tail dependence obtained by endowing the support points in de Haan's spectral representation of max-stable processes with velocities and lifetimes. We extend Smith's model to these max-stable velocity processes and obtain the distribution of waiting times between extreme events at multiple locations. Motivated by this result, a new definition of tail dependence is proposed that is a function of the distribution of waiting times between threshold exceedances, and an inferential framework is constructed for estimating the strength of extremal dependence and quantifying uncertainty in this paradigm. The method is applied to climatological, financial, and electrophysiology data.

The remainder of this thesis focuses on posterior computation by Markov chain Monte Carlo. The Markov Chain Monte Carlo method is the dominant paradigm for posterior computation in Bayesian analysis. It has long been common to control computation time by making approximations to the Markov transition kernel. Comparatively little attention has been paid to convergence and estimation error in these approximating Markov Chains. In Chapter 6, we propose a framework for assessing when to use approximations in MCMC algorithms, and how much error in the transition kernel should be tolerated to obtain optimal estimation performance with respect to a specified loss function and computational budget. The results require only ergodicity of the exact kernel and control of the kernel approximation accuracy. The theoretical framework is applied to approximations based on random subsets of data, low-rank approximations of Gaussian processes, and a novel approximating Markov chain for discrete mixture models.

Data augmentation Gibbs samplers are arguably the most popular class of algorithm for approximately sampling from the posterior distribution for the parameters of generalized linear models. The truncated Normal and Polya-Gamma data augmentation samplers are standard examples for probit and logit links, respectively. Motivated by an important problem in quantitative advertising, in Chapter 7 we consider the application of these algorithms to modeling rare events. We show that when the sample size is large but the observed number of successes is small, these data augmentation samplers mix very slowly, with a spectral gap that converges to zero at a rate at least proportional to the reciprocal of the square root of the sample size up to a log factor. In simulation studies, moderate sample sizes result in high autocorrelations and small effective sample sizes. Similar empirical results are observed for related data augmentation samplers for multinomial logit and probit models. When applied to a real quantitative advertising dataset, the data augmentation samplers mix very poorly. Conversely, Hamiltonian Monte Carlo and a type of independence chain Metropolis algorithm show good mixing on the same dataset.

Integrating network analysis, sensor tags, and observation to understand shark ecology and behavior

Group living in animals is a well-studied phenomenon, having been documented extensively in a wide range of terrestrial, freshwater, and marine species. Although social dynamics are complex across space and time, recent technological and analytical advances enable deeper understanding of their nature and ecological implications. While for some taxa, a great deal of information is known regarding the mechanistic underpinnings of these social processes, knowledge of these mechanisms in elasmobranchs is lacking. Here, we used an integrative and novel combination of direct observation, accelerometer biologgers, and recent advances in network analysis to better understand the mechanistic bases of individual-level differences in sociality (leadership, network attributes) and diel patterns of locomotor activity in a widespread marine predator, the lemon shark (Negaprion brevirostris). We found that dynamic models of interaction based on Markov chains can accurately predict juvenile lemon shark social behavior and that lemon sharks did not occupy consistent positions within their network. Lemon sharks did however preferentially associate with specific group members, by sex as well as by similarity or nonsimilarity for a number of behavioral (nonsimilarity: leadership) and locomotor traits (similarity: proportion of time swimming "fast," mean swim duration; nonsimilarity: proportion of swimming bursts/transitions between activity states). Our study provides some of the first information on the mechanistic bases of group living and personality in sharks and further, a potential experimental approach for studying fine-scale differences in behavior and locomotor patterns in difficult-to-study organisms.

Call recognition and individual identification of fish vocalizations based on automatic speech recognition: An example with the Lusitanian toadfish

The study of acoustic communication in animals often requires not only the recognition of species specific acoustic signals but also the identification of individual subjects, all in a complex acoustic background. Moreover, when very long recordings are to be analyzed, automatic recognition and identification processes are invaluable tools to extract the relevant biological information. A pattern recognition methodology based on hidden Markov models is presented inspired by successful results obtained in the most widely known and complex acoustical communication signal: human speech. This methodology was applied here for the first time to the detection and recognition of fish acoustic signals, specifically in a stream of round-the-clock recordings of Lusitanian toadfish (Halobatrachus didactylus) in their natural estuarine habitat. The results show that this methodology is able not only to detect the mating sounds (boatwhistles) but also to identify individual male toadfish, reaching an identification rate of ca. 95%. Moreover this method also proved to be a powerful tool to assess signal durations in large data sets. However, the system failed in recognizing other sound types.

Assuring virtual network reliability and resilience

A framework developed that uses reliability block diagrams and continuous-time Markov chains to model and analyse the reliability and availability of a Virtual Network Environment (VNE). In addition, to minimize the unpredicted failures and reduce the impact of failure on a virtual network, a dynamic solution proposed for detecting a failure before it occurs in the VNE. Moreover, to predict failure and establish a tolerable maintenance plan before failure occurs in the VNE, a failure prediction method for VNE can be used to minimise the unpredicted failures, reduce backup redundancy and maximise system performance.

Convergence time to equilibrium distributions of autonomous and periodic non-autonomous graphs

We present some estimates of the time of convergence to the equilibrium distribution in autonomous and periodic non-autonomous graphs, with ergodic stochastic adjacency matrices, using the eigenvalues of these matrices. On this way we generalize previous results from several authors, that only considered reversible matrices.