A principled approach to interactive hierarchical non-linear visualization of high-dimensional data

Hierarchical visualization systems are desirable because a single two-dimensional visualization plot may not be sufficient to capture all of the interesting aspects of complex high-dimensional data sets. We extend an existing locally linear hierarchical visualization system PhiVis [1] in several directions: bf(1) we allow for em non-linear projection manifolds (the basic building block is the Generative Topographic Mapping -- GTM), bf(2) we introduce a general formulation of hierarchical probabilistic models consisting of local probabilistic models organized in a hierarchical tree, bf(3) we describe folding patterns of low-dimensional projection manifold in high-dimensional data space by computing and visualizing the manifold's local directional curvatures. Quantities such as magnification factors [3] and directional curvatures are helpful for understanding the layout of the nonlinear projection manifold in the data space and for further refinement of the hierarchical visualization plot. Like PhiVis, our system is statistically principled and is built interactively in a top-down fashion using the EM algorithm. We demonstrate the visualization system principle of the approach on a complex 12-dimensional data set and mention possible applications in the pharmaceutical industry.

Constructing localized non-linear projection manifolds in a principled way:hierarchical generative topographic mapping

It has been argued that a single two-dimensional visualization plot may not be sufficient to capture all of the interesting aspects of complex data sets, and therefore a hierarchical visualization system is desirable. In this paper we extend an existing locally linear hierarchical visualization system PhiVis (Bishop98a) in several directions: 1. We allow for em non-linear projection manifolds. The basic building block is the Generative Topographic Mapping. 2. We introduce a general formulation of hierarchical probabilistic models consisting of local probabilistic models organized in a hierarchical tree. General training equations are derived, regardless of the position of the model in the tree. 3. Using tools from differential geometry we derive expressions for local directionalcurvatures of the projection manifold. Like PhiVis, our system is statistically principled and is built interactively in a top-down fashion using the EM algorithm. It enables the user to interactively highlight those data in the parent visualization plot which are captured by a child model.We also incorporate into our system a hierarchical, locally selective representation of magnification factors and directional curvatures of the projection manifolds. Such information is important for further refinement of the hierarchical visualization plot, as well as for controlling the amount of regularization imposed on the local models. We demonstrate the principle of the approach on a toy data set andapply our system to two more complex 12- and 19-dimensional data sets.

Wind power forecasts using Gaussian processes and numerical weather prediction

Since wind at the earth's surface has an intrinsically complex and stochastic nature, accurate wind power forecasts are necessary for the safe and economic use of wind energy. In this paper, we investigated a combination of numeric and probabilistic models: a Gaussian process (GP) combined with a numerical weather prediction (NWP) model was applied to wind-power forecasting up to one day ahead. First, the wind-speed data from NWP was corrected by a GP, then, as there is always a defined limit on power generated in a wind turbine due to the turbine controlling strategy, wind power forecasts were realized by modeling the relationship between the corrected wind speed and power output using a censored GP. To validate the proposed approach, three real-world datasets were used for model training and testing. The empirical results were compared with several classical wind forecast models, and based on the mean absolute error (MAE), the proposed model provides around 9% to 14% improvement in forecasting accuracy compared to an artificial neural network (ANN) model, and nearly 17% improvement on a third dataset which is from a newly-built wind farm for which there is a limited amount of training data. © 2013 IEEE.

Построение Логико-Вероятностных Моделей Временных Рядов с Использованием Цепей Маркова

The method of logic and probabilistic models constructing for multivariate heterogeneous time series is offered. There are some important properties of these models, e.g. universality. In this paper also discussed the logic and probabilistic models distinctive features in comparison with hidden Markov processes. The early proposed time series forecasting algorithm is tested on applied task.

Short-term wind power forecasting using Gaussian processes

Since wind has an intrinsically complex and stochastic nature, accurate wind power forecasts are necessary for the safety and economics of wind energy utilization. In this paper, we investigate a combination of numeric and probabilistic models: one-day-ahead wind power forecasts were made with Gaussian Processes (GPs) applied to the outputs of a Numerical Weather Prediction (NWP) model. Firstly the wind speed data from NWP was corrected by a GP. Then, as there is always a defined limit on power generated in a wind turbine due the turbine controlling strategy, a Censored GP was used to model the relationship between the corrected wind speed and power output. To validate the proposed approach, two real world datasets were used for model construction and testing. The simulation results were compared with the persistence method and Artificial Neural Networks (ANNs); the proposed model achieves about 11% improvement in forecasting accuracy (Mean Absolute Error) compared to the ANN model on one dataset, and nearly 5% improvement on another.

Immunological synapse: a mathematical model of the bond formation process:mathematical modelling of the immature synapse

The cell:cell bond between an immune cell and an antigen presenting cell is a necessary event in the activation of the adaptive immune response. At the juncture between the cells, cell surface molecules on the opposing cells form non-covalent bonds and a distinct patterning is observed that is termed the immunological synapse. An important binding molecule in the synapse is the T-cell receptor (TCR), that is responsible for antigen recognition through its binding with a major-histocompatibility complex with bound peptide (pMHC). This bond leads to intracellular signalling events that culminate in the activation of the T-cell, and ultimately leads to the expression of the immune eector function. The temporal analysis of the TCR bonds during the formation of the immunological synapse presents a problem to biologists, due to the spatio-temporal scales (nanometers and picoseconds) that compare with experimental uncertainty limits. In this study, a linear stochastic model, derived from a nonlinear model of the synapse, is used to analyse the temporal dynamics of the bond attachments for the TCR. Mathematical analysis and numerical methods are employed to analyse the qualitative dynamics of the nonequilibrium membrane dynamics, with the specic aim of calculating the average persistence time for the TCR:pMHC bond. A single-threshold method, that has been previously used to successfully calculate the TCR:pMHC contact path sizes in the synapse, is applied to produce results for the average contact times of the TCR:pMHC bonds. This method is extended through the development of a two-threshold method, that produces results suggesting the average time persistence for the TCR:pMHC bond is in the order of 2-4 seconds, values that agree with experimental evidence for TCR signalling. The study reveals two distinct scaling regimes in the time persistent survival probability density prole of these bonds, one dominated by thermal uctuations and the other associated with the TCR signalling. Analysis of the thermal fluctuation regime reveals a minimal contribution to the average time persistence calculation, that has an important biological implication when comparing the probabilistic models to experimental evidence. In cases where only a few statistics can be gathered from experimental conditions, the results are unlikely to match the probabilistic predictions. The results also identify a rescaling relationship between the thermal noise and the bond length, suggesting a recalibration of the experimental conditions, to adhere to this scaling relationship, will enable biologists to identify the start of the signalling regime for previously unobserved receptor:ligand bonds. Also, the regime associated with TCR signalling exhibits a universal decay rate for the persistence probability, that is independent of the bond length.

Bayesian Learning with Dependency Structures via Latent Factors, Mixtures, and Copulas

Bayesian methods offer a flexible and convenient probabilistic learning framework to extract interpretable knowledge from complex and structured data. Such methods can characterize dependencies among multiple levels of hidden variables and share statistical strength across heterogeneous sources. In the first part of this dissertation, we develop two dependent variational inference methods for full posterior approximation in non-conjugate Bayesian models through hierarchical mixture- and copula-based variational proposals, respectively. The proposed methods move beyond the widely used factorized approximation to the posterior and provide generic applicability to a broad class of probabilistic models with minimal model-specific derivations. In the second part of this dissertation, we design probabilistic graphical models to accommodate multimodal data, describe dynamical behaviors and account for task heterogeneity. In particular, the sparse latent factor model is able to reveal common low-dimensional structures from high-dimensional data. We demonstrate the effectiveness of the proposed statistical learning methods on both synthetic and real-world data.

Contextual text mining

With the dramatic growth of text information, there is an increasing need for powerful text mining systems that can automatically discover useful knowledge from text. Text is generally associated with all kinds of contextual information. Those contexts can be explicit, such as the time and the location where a blog article is written, and the author(s) of a biomedical publication, or implicit, such as the positive or negative sentiment that an author had when she wrote a product review; there may also be complex context such as the social network of the authors. Many applications require analysis of topic patterns over different contexts. For instance, analysis of search logs in the context of the user can reveal how we can improve the quality of a search engine by optimizing the search results according to particular users; analysis of customer reviews in the context of positive and negative sentiments can help the user summarize public opinions about a product; analysis of blogs or scientific publications in the context of a social network can facilitate discovery of more meaningful topical communities. Since context information significantly affects the choices of topics and language made by authors, in general, it is very important to incorporate it into analyzing and mining text data. In general, modeling the context in text, discovering contextual patterns of language units and topics from text, a general task which we refer to as Contextual Text Mining, has widespread applications in text mining. In this thesis, we provide a novel and systematic study of contextual text mining, which is a new paradigm of text mining treating context information as the ``first-class citizen.'' We formally define the problem of contextual text mining and its basic tasks, and propose a general framework for contextual text mining based on generative modeling of text. This conceptual framework provides general guidance on text mining problems with context information and can be instantiated into many real tasks, including the general problem of contextual topic analysis. We formally present a functional framework for contextual topic analysis, with a general contextual topic model and its various versions, which can effectively solve the text mining problems in a lot of real world applications. We further introduce general components of contextual topic analysis, by adding priors to contextual topic models to incorporate prior knowledge, regularizing contextual topic models with dependency structure of context, and postprocessing contextual patterns to extract refined patterns. The refinements on the general contextual topic model naturally lead to a variety of probabilistic models which incorporate different types of context and various assumptions and constraints. These special versions of the contextual topic model are proved effective in a variety of real applications involving topics and explicit contexts, implicit contexts, and complex contexts. We then introduce a postprocessing procedure for contextual patterns, by generating meaningful labels for multinomial context models. This method provides a general way to interpret text mining results for real users. By applying contextual text mining in the ``context'' of other text information management tasks, including ad hoc text retrieval and web search, we further prove the effectiveness of contextual text mining techniques in a quantitative way with large scale datasets. The framework of contextual text mining not only unifies many explorations of text analysis with context information, but also opens up many new possibilities for future research directions in text mining.

Online MBPeT dashboard : enabling performance testing as a service in the cloud

In this thesis, tool support is addressed for the combined disciplines of Model-based testing and performance testing. Model-based testing (MBT) utilizes abstract behavioral models to automate test generation, thus decreasing time and cost of test creation. MBT is a functional testing technique, thereby focusing on output, behavior, and functionality. Performance testing, however, is non-functional and is concerned with responsiveness and stability under various load conditions. MBPeT (Model-Based Performance evaluation Tool) is one such tool which utilizes probabilistic models, representing dynamic real-world user behavior patterns, to generate synthetic workload against a System Under Test and in turn carry out performance analysis based on key performance indicators (KPI). Developed at Åbo Akademi University, the MBPeT tool is currently comprised of a downloadable command-line based tool as well as a graphical user interface. The goal of this thesis project is two-fold: 1) to extend the existing MBPeT tool by deploying it as a web-based application, thereby removing the requirement of local installation, and 2) to design a user interface for this web application which will add new user interaction paradigms to the existing feature set of the tool. All phases of the MBPeT process will be realized via this single web deployment location including probabilistic model creation, test configurations, test session execution against a SUT with real-time monitoring of user configurable metric, and final test report generation and display. This web application (MBPeT Dashboard) is implemented with the Java programming language on top of the Vaadin framework for rich internet application development. The Vaadin framework handles the complicated web communications processes and front-end technologies, freeing developers to implement the business logic as well as the user interface in pure Java. A number of experiments are run in a case study environment to validate the functionality of the newly developed Dashboard application as well as the scalability of the solution implemented in handling multiple concurrent users. The results support a successful solution with regards to the functional and performance criteria defined, while improvements and optimizations are suggested to increase both of these factors.

Combined exposure to multiple mycotoxins in food: are we ready for risk assessment and risk management?

It is nowadays recognized that the risk of human co-exposure to multiple mycotoxins is real. In the last years, a number of studies have approached the issue of co-exposure and the best way to develop a more precise and realistic assessment. Likewise, the growing concern about the combined effects of mycotoxins and their potential impact on human health has been reflected by the increasing number of toxicological studies on the combined toxicity of these compounds. Nevertheless, risk assessment of these toxins, still follows the conventional paradigm of single exposure and single effects, incorporating only the possibility of additivity but not taking into account the complex dynamics associated to interactions between different mycotoxins or between mycotoxins and other food contaminants. Considering that risk assessment is intimately related to the establishment of regulatory guidelines, once the risk assessment is completed, an effort to reduce or manage the risk should be followed to protect public health. Risk assessment of combined human exposure to multiple mycotoxins thus poses several challenges to scientists, risk assessors and risk managers and opens new avenues for research. This presentation aims to give an overview of the different challenges posed by the likelihood of human co-exposure to mycotoxins and the possibility of interactive effects occurring after absorption, towards knowledge generation to support a more accurate human risk assessment and risk management. For this purpose, a physiologically-based framework that includes knowledge on the bioaccessibility, toxicokinetics and toxicodynamics of multiple toxins is proposed. Regarding exposure assessment, the need of harmonized food consumption data, availability of multianalyte methods for mycotoxin quantification, management of left-censored data and use of probabilistic models will be highlight, in order to develop a more precise and realistic exposure assessment. On the other hand, the application of predictive mathematical models to estimate mycotoxins’ combined effects from in vitro toxicity studies will be also discussed. Results from a recent Portuguese project aimed at exploring the toxic effects of mixtures of mycotoxins in infant foods and their potential health impact will be presented as a case study, illustrating the different aspects of risk assessment highlighted in this presentation. Further studies on hazard and exposure assessment of multiple mycotoxins, using harmonized approaches and methodologies, will be crucial towards an improvement in data quality and contributing to holistic risk assessment and risk management strategies for multiple mycotoxins in foodstuffs.

BOA for nurse scheduling

Our research has shown that schedules can be built mimicking a human scheduler by using a set of rules that involve domain knowledge. This chapter presents a Bayesian Optimization Algorithm (BOA)for the nurse scheduling problem that chooses such suitable scheduling rules from a set for each nurse’s assignment. Based on the idea of using probabilistic models, the BOA builds a Bayesian network for the set of promising solutions and samples these networks to generate new candidate solutions. Computational results from 52 real data instances demonstrate the success of this approach. It is also suggested that the learning mechanism in the proposed algorithm may be suitable for other scheduling problems.

SAMPLING BIAS IN EVALUATING THE PROBABILITY OF SEISMICALLY INDUCED SOIL LIQUEFACTION WITH SPT & CPT CASE HISTORIES

Several deterministic and probabilistic methods are used to evaluate the probability of seismically induced liquefaction of a soil. The probabilistic models usually possess some uncertainty in that model and uncertainties in the parameters used to develop that model. These model uncertainties vary from one statistical model to another. Most of the model uncertainties are epistemic, and can be addressed through appropriate knowledge of the statistical model. One such epistemic model uncertainty in evaluating liquefaction potential using a probabilistic model such as logistic regression is sampling bias. Sampling bias is the difference between the class distribution in the sample used for developing the statistical model and the true population distribution of liquefaction and non-liquefaction instances. Recent studies have shown that sampling bias can significantly affect the predicted probability using a statistical model. To address this epistemic uncertainty, a new approach was developed for evaluating the probability of seismically-induced soil liquefaction, in which a logistic regression model in combination with Hosmer-Lemeshow statistic was used. This approach was used to estimate the population (true) distribution of liquefaction to non-liquefaction instances of standard penetration test (SPT) and cone penetration test (CPT) based most updated case histories. Apart from this, other model uncertainties such as distribution of explanatory variables and significance of explanatory variables were also addressed using KS test and Wald statistic respectively. Moreover, based on estimated population distribution, logistic regression equations were proposed to calculate the probability of liquefaction for both SPT and CPT based case history. Additionally, the proposed probability curves were compared with existing probability curves based on SPT and CPT case histories.

Analysis of earthquakes predictability

Earthquake prediction is a complex task for scientists due to the rare occurrence of high-intensity earthquakes and their inaccessible depths. Despite this challenge, it is a priority to protect infrastructure, and populations living in areas of high seismic risk. Reliable forecasting requires comprehensive knowledge of seismic phenomena. In this thesis, the development, application, and comparison of both deterministic and probabilistic forecasting methods is shown. Regarding the deterministic approach, the implementation of an alarm-based method using the occurrence of strong (fore)shocks, widely felt by the population, as a precursor signal is described. This model is then applied for retrospective prediction of Italian earthquakes of magnitude M≥5.0,5.5,6.0, occurred in Italy from 1960 to 2020. Retrospective performance testing is carried out using tests and statistics specific to deterministic alarm-based models. Regarding probabilistic models, this thesis focuses mainly on the EEPAS and ETAS models. Although the EEPAS model has been previously applied and tested in some regions of the world, it has never been used for forecasting Italian earthquakes. In the thesis, the EEPAS model is used to retrospectively forecast Italian shallow earthquakes with a magnitude of M≥5.0 using new MATLAB software. The forecasting performance of the probabilistic models was compared to other models using CSEP binary tests. The EEPAS and ETAS models showed different characteristics for forecasting Italian earthquakes, with EEPAS performing better in the long-term and ETAS performing better in the short-term. The FORE model based on strong precursor quakes is compared to EEPAS and ETAS using an alarm-based deterministic approach. All models perform better than a random forecasting model, with ETAS and FORE models showing better performance. However, to fully evaluate forecasting performance, prospective tests should be conducted. The lack of objective tests for evaluating deterministic models and comparing them with probabilistic ones was a challenge faced during the study.

Towards a logical foundation of randomized computation

This dissertation investigates the relations between logic and TCS in the probabilistic setting. It is motivated by two main considerations. On the one hand, since their appearance in the 1960s-1970s, probabilistic models have become increasingly pervasive in several fast-growing areas of CS. On the other, the study and development of (deterministic) computational models has considerably benefitted from the mutual interchanges between logic and CS. Nevertheless, probabilistic computation was only marginally touched by such fruitful interactions. The goal of this thesis is precisely to (start) bring(ing) this gap, by developing logical systems corresponding to specific aspects of randomized computation and, therefore, by generalizing standard achievements to the probabilistic realm. To do so, our key ingredient is the introduction of new, measure-sensitive quantifiers associated with quantitative interpretations. The dissertation is tripartite. In the first part, we focus on the relation between logic and counting complexity classes. We show that, due to our classical counting propositional logic, it is possible to generalize to counting classes, the standard results by Cook and Meyer and Stockmeyer linking propositional logic and the polynomial hierarchy. Indeed, we show that the validity problem for counting-quantified formulae captures the corresponding level in Wagner's hierarchy. In the second part, we consider programming language theory. Type systems for randomized \lambda-calculi, also guaranteeing various forms of termination properties, were introduced in the last decades, but these are not "logically oriented" and no Curry-Howard correspondence is known for them. Following intuitions coming from counting logics, we define the first probabilistic version of the correspondence. Finally, we consider the relationship between arithmetic and computation. We present a quantitative extension of the language of arithmetic able to formalize basic results from probability theory. This language is also our starting point to define randomized bounded theories and, so, to generalize canonical results by Buss.