In this thesis, an approach for localizing culprits of degradation of quality measures in an IPTV streaming service using Bayesian net-work is presented. This task is referred to as Root Cause Analysis(RCA). The objective of this thesis is to develop a model that is able to provide useful information to technicians by generating a list of probable root causes in order to shorten the amount of time spent on trouble shooting. A performance comparison is presented in Section Experimental results with Bayesian models such as Naive Bayes (NB),Tree Augmented naive Bayes (TAN) and Hill Climbing (HC) and the non Bayesian methods K-Nearest Neighbors and Random Forest. The results of the RCA models indicated that the most frequent most prob-able cause of degradation of quality is the signal strength of the user’s Wi-Fi that is reported at the user’s TV box. / I detta examensarbete presenteras en metod för att lokalisera grundorsaken till nedgradering av kvalitet i en IPTV strömningstjänst. Denna uppgift refererar tillgrundorsaksanalys. Avsikten med denna tes är att utveckla en modell som kan tillförse tekniker med användarbar information genom att generera en lista med möjliga grundorsaker för att förkorta tiden som spenderas med felsökning. En prestandajämförelse är presenterad i Sektion Experimental results med de Bayesianska modellerna Naive Bayes (NB), Tree Augmented naive Bayes (TAN) och Hill Climbing (HC) samt de icke Bayesianska modellerna K-Nearest Neighbors och Random Forest. Resultatet av grundorsaksmodellerna indikerade att den mest frekventa mest sannolika grundorsaken till nedgradering av kvalitet är signal styrkan hos Wi-Fi nätverket vilket rapporteras i användarens TV-box.

This study investigates E-banking Operational Risk Assessment (ORA) to enable the development of a new ORA framework and methodology. The general view is that E-banking systems have modified some of the traditional banking risks, particularly Operational Risk (OR) as suggested by the Basel Committee on Banking Supervision in 2003. In addition, recent E-banking financial losses together with risk management principles and standards raise the need for an effective ORA methodology and framework in the context of E-banking. Moreover, evaluation tools and / or methods for ORA are highly subjective, are still in their infant stages, and have not yet reached a consensus. Therefore, it is essential to develop valid and reliable methods for effective ORA and evaluations. The main contribution of this thesis is to apply Fuzzy Inference System (FIS) and Tree Augmented Naïve Bayes (TAN) classifier as standard tools for identifying OR, and measuring OR exposure level. In addition, a new ORA methodology is proposed which consists of four major steps: a risk model, assessment approach, analysis approach and a risk assessment process. Further, a new ORA framework and measurement metrics are proposed with six factors: frequency of triggering event, effectiveness of avoidance barriers, frequency of undesirable operational state, effectiveness of recovery barriers before the risk outcome, approximate cost for Undesirable Operational State (UOS) occurrence, and severity of the risk outcome. The study results were reported based on surveys conducted with Nigerian senior banking officers and banking customers. The study revealed that the framework and assessment tools gave good predictions for risk learning and inference in such systems. Thus, results obtained can be considered promising and useful for both E-banking system adopters and future researchers in this area.

Probabilistic graphical models like Bayesian Networks (BNs) are powerful artificial-intelligence formalisms, with similarities to cognition and higher order reasoning in the human brain. These models have been, to great success, applied to several challenging real-world applications. Use of these formalisms to a greater set of applications is impeded by the limitations of the currently used software-based implementations. New emerging-technology based circuit paradigms which leverage physical equivalence, i.e., operating directly on probabilities vs. introducing layers of abstraction, promise orders of magnitude increase in performance and efficiency of BN implementations, enabling networks with millions of random variables. While majority of applications with small network size (100s of nodes) require only single digit precision for accurate results, applications with larger size (1000s to millions of nodes) require higher precision computation. We introduce a new BN integrated circuit fabric based on mixed-signal magneto-electric circuits which perform probabilistic computations based on the principle of approximate computation. Precision scaling in this fabric is logarithmic in area vs. linear in prior directions. Results show 33x area benefit for a 0.001 precision compared to prior direction, while maintaining three orders of magnitude performance benefits vs. 100-core processor implementations.

The increase in diagnostic and therapeutic procedures in the treatment of oncological diseases, as well as the limited capacity of experts to provide information, necessitates the development of therapy decision support systems (TDSS). We have developed a treatment decision model that integrates available patient information as well as tumor characteristics. They are assessed according to their relevance in evaluating the optimal therapy option. Our treatment model is based on Bayesian networks (BN) which integrate patient-specific data with expert-based implemented causalities to suggest the optimal therapy option and therefore potentially support the decision-making process for treatment of laryngeal carcinoma. To test the reliability of our model, we compared the calculations of our model with the documented therapy from our data set, which contained information on 97 patients with laryngeal carcinoma. Information on 92 patients was used in our analyses and the model suggested the correct treatment in 419 out of 460 treatment modalities (accuracy of 91%). However, unequally distributed clinical data in the test sets revealed weak spots in the model that require revision for future utilization.

The piping plover (Charadrius melodus) is a federally-listed shorebird that nests on barrier islands along the U.S. Atlantic Coast and is highly vulnerable to habitat change and predation. We have addressed these two threats by 1) developing and implementing a linked model system that predicts future change to piping plover habitat resulting from sea-level rise and beach management efforts by joining dynamic models of sea-level rise, shoreline change, island geomorphology and piping plover nest habitat suitability, and 2) quantifying occupancy and movement of the red fox (Vulpes vulpes), a key shorebird predator at Assateague Island, Maryland and Virginia. We constructed and tested a model that links changes in geomorphological characteristics to piping plover nesting habitat suitability. We then linked this model to larger scale shoreline change resulting from sea level rise and storms. Using this linked model to forecast future sea-level rise and beach management efforts, we found that modest sea-level rise rates (3 mm and 4.1 mm/yr; similar to current rates) may increase suitable piping plover nesting habitat area in 50-100 years and some beach management strategies (beach nourishment and artificial dune modifications) also influence habitat availability. Our development and implementation of this tool to predict change in piping plover habitat suitability provides a vital starting point for predicting how plover nesting habitat will change in a context of planned human modifications intended to address climate change-related threats. Our findings regarding red fox occupancy and movement complement the use of this model for planning future management actions by providing vital information on the effects of certain predator management activities and habitat use of a key mammalian predator, the red fox, for shorebirds along the U.S. Atlantic Coast. Overall, we found that 1) red fox occupancy was strongly tied to eastern cottontail (Sylvilagus floridanus) trap success, increasing sharply with increased eastern cottontail trap success, 2) red fox occupancy did not change in response to an intensive eradication program, and 3) red foxes in our study area generally moved little between camera stations spaced 300 m from each other, but may move large distances (> 6km) at times, likely to occupy new territory available after lethal control efforts. Our findings have important ramifications for the sustainability of long-term predator removal programs and our understanding of future habitat change on the red fox. For example how vegetation changes affect eastern cottontails, how resulting fluctuations in eastern cottontails affect red fox occupancy, and how consequential changes in red fox occupancy affect plover breeding productivity. Our predictive model combined with these predator findings will allow wildlife managers to better plan and implement effective management actions for piping plovers in response to the multiple stressors of SLR-induced habitat change and predation. / Ph. D.

Keyword analysis in scientific articles is a method used to identify and evaluate the importance and relevance of specific words or phrases (keywords) within scientific literature. The primary goal of keyword analysis is to uncover the core themes, research trends, and conceptual frameworks within a given field or across multiple disciplines. It helps researchers understand scientific discourse's focus and ideas' evolution over time. This thesis performs keyword analysis on a repository of scientific publications through a combination of methods. It starts with extracting the available keywords, and it deals with the missing keywords data through data augmentation. Then, it utilizes a variety of statistical methods to gain insight into the publications. The study employs an implementation of LDA topic modeling to accurately categorize keywords into thematic groups, a Vector autoregression to explore keyword relationships, and temporal dynamics of keywords. Next, the research further examines the interdisciplinary connectivity of keywords, clarifying the collective nature of modern science. In conclusion, the thesis presents a comprehensive framework for keyword analysis in scientific literature, through a blend of data augmentation, natural language processing, temporal dynamics, and interdisciplinary examination, the study provides a robust tool for understanding the development and structure of scientific literature. The findings of this research have important implications for scholars, it allows navigating the vast amount of scientific literature more effectively and to discern the most influential ideas and trends shaping target fields. The methodologies implemented here offer an opportunity for any studies to methodologically search, extract, and identify keywords to find relevant papers and interpret the complex landscape of scientific communication. / Nyckelordsanalys i vetenskapliga artiklar är en metod som används för att identifiera och utvärdera vikten och relevansen av specifika ord eller fraser (nyckelord) inom vetenskaplig litteratur. Det primära målet med nyckelordsanalys är att avslöja kärnteman, forskningstrender och konceptuella ramverk inom ett givet fält eller över flera lämnar. Det hjälper forskare att förstå den vetenskapliga diskursens fokus och idéernas utveckling över tid. Denna avhandling utför nyckelordsanalys på ett arkiv av vetenskapliga publikationer genom en kombination av metoder. Den börjar med att extrahera de tillgängliga nyckelorden och hanterar de saknade nyckelordsdata genom dataaugtation. Därefter använder den en mängd statistiska metoder för att få insikt i publikationerna. Studien använder en implementering av LDA-ämnesmodellering för att noggrant kategorisera nyckelord i tematiska grupper, en vektorautoregression för att utforska nyckelordsrelationer och tidsmässig dynamik av nyckelord. Nästa steg i forskningen är att ytterligare undersöka den tvärvetenskapliga kopplingen mellan nyckelord, vilket klargör den kollektiva naturen av modern vetenskap. Sammanfattningsvis presenterar avhandlingen ett omfattande ramverk för nyckelordsanalys i vetenskaplig litteratur. Genom en blandning av dataaugmentation, naturlig språkbehandling, tidsmässig dynamik och tvärvetenskaplig undersökning, erbjuder studien ett robust verktyg för att förstå utvecklingen och strukturen av vetenskaplig litteratur. Forskningens resultat har viktiga implikationer för forskare; det möjliggör effektivare navigering i den omfattande mängden vetenskaplig litteratur och att urskilja de mest inflytdrikelserika idéerna och trenderna som formar målfälten. De metoder som införas här erbjuder en möjlighet för vilken studie som helst att metodiskt söka, extrahera och identifiera nyckelord för att hitta relevanta artiklar och tolka det komplexa landskapet av vetenskaplig kommunikation.

Estimation of Distribution Algorithms (EDAs) belong to the class of population based optimisation algorithms. They are motivated by the idea of discovering and exploiting the interaction between variables in the solution. They estimate a probability distribution from population of solutions, and sample it to generate the next population. Many EDAs use probabilistic graphical modelling techniques for this purpose. In particular, directed graphical models (Bayesian networks) have been widely used in EDA. This thesis proposes an undirected graphical model (Markov Random Field (MRF)) approach to estimate and sample the distribution in EDAs. The interaction between variables in the solution is modelled as an undirected graph and the joint probability of a solution is factorised as a Gibbs distribution. The thesis describes a model of fitness function that approximates the energy in the Gibbs distribution, and shows how this model can be fitted to a population of solutions to estimate the parameters of the MRF. The estimated MRF is then sampled to generate the next population. This approach is applied to estimation of distribution in a general framework of an EDA, called Distribution Estimation using Markov Random Fields (DEUM). The thesis then proposes several variants of DEUM using different sampling techniques and tests their performance on a range of optimisation problems. The results show that, for most of the tested problems, the DEUM algorithms significantly outperform other EDAs, both in terms of number of fitness evaluations and the quality of the solutions found by them. There are two main explanations for the success of DEUM algorithms. Firstly, DEUM builds a model of fitness function to approximate the MRF. This contrasts with other EDAs, which build a model of selected solutions. This allows DEUM to use fitness in variation part of the evolution. Secondly, DEUM exploits the temperature coefficient in the Gibbs distribution to regulate the behaviour of the algorithm. In particular, with higher temperature, the distribution is closer to being uniform and with lower temperature it concentrates near some global optima. This gives DEUM an explicit control over the convergence of the algorithm, resulting in better optimisation.

Nous proposons une approche probabiliste afin de déterminer l’impact des changements dans les programmes à objets. Cette approche sert à prédire, pour un changement donné dans une classe du système, l’ensemble des autres classes potentiellement affectées par ce changement. Cette prédiction est donnée sous la forme d’une probabilité qui dépend d’une part, des interactions entre les classes exprimées en termes de nombre d’invocations et d’autre part, des relations extraites à partir du code source. Ces relations sont extraites automatiquement par rétro-ingénierie. Pour la mise en oeuvre de notre approche, nous proposons une approche basée sur les réseaux bayésiens. Après une phase d’apprentissage, ces réseaux prédisent l’ensemble des classes affectées par un changement. L’approche probabiliste proposée est évaluée avec deux scénarios distincts mettant en oeuvre plusieurs types de changements effectués sur différents systèmes. Pour les systèmes qui possèdent des données historiques, l’apprentissage a été réalisé à partir des anciennes versions. Pour les systèmes dont on ne possède pas assez de données relatives aux changements de ses versions antécédentes, l’apprentissage a été réalisé à l’aide des données extraites d’autres systèmes. / We study the possibility of predicting the impact of changes in object-oriented code using bayesian networks. For each change type, we produce a bayesian network that determines the probability that a class is impacted given that another class is changed. Each network takes as input a set of possible relationships between classes. We train our networks using historical data. The proposed impact-prediction approach is evaluated with two different scenarios, various types of changes, and five systems. In the first scenario, we use as training data, the changes performed in the previous versions of the same system. In the second scenario training data is borrowed from systems that are different from the changed one. Our evaluation showed that, in both cases, we obtain very good predictions, even though they are better in the first scenario.

The research work presented in this thesis describes the development and testing of a new data analysis methodology for the automated near real-time detection and approximate location of pipe bursts and other events which induce similar abnormal pressure/flow variations (e.g., unauthorised consumptions, equipment failures, etc.) in Water Distribution Systems (WDSs). This methodology makes synergistic use of several self-learning Artificial Intelligence (AI) and statistical/geostatistical techniques for the analysis of the stream of data (i.e., signals) collected and communicated on-line by the hydraulic sensors deployed in a WDS. These techniques include: (i) wavelets for the de-noising of the recorded pressure/flow signals, (ii) Artificial Neural Networks (ANNs) for the short-term forecasting of future pressure/flow signal values, (iii) Evolutionary Algorithms (EAs) for the selection of optimal ANN input structure and parameters sets, (iv) Statistical Process Control (SPC) techniques for the short and long term analysis of the burst/other event-induced pressure/flow variations, (v) Bayesian Inference Systems (BISs) for inferring the probability of a burst/other event occurrence and raising the detection alarms, and (vi) geostatistical techniques for determining the approximate location of a detected burst/other event. The results of applying the new methodology to the pressure/flow data from several District Metered Areas (DMAs) in the United Kingdom (UK) with real-life bursts/other events and simulated (i.e., engineered) burst events are also reported in this thesis. The results obtained illustrate that the developed methodology allowed detecting the aforementioned events in a fast and reliable manner and also successfully determining their approximate location within a DMA. The results obtained additionally show the potential of the methodology presented here to yield substantial improvements to the state-of-the-art in near real-time WDS incident management by enabling the water companies to save water, energy, money, achieve higher levels of operational efficiency and improve their customer service. The new data analysis methodology developed and tested as part of the research work presented in this thesis has been patented (International Application Number: PCT/GB2010/000961).

This study investigates E-banking Operational Risk Assessment (ORA) to enable the development of a new ORA framework and methodology. The general view is that E-banking systems have modified some of the traditional banking risks, particularly Operational Risk (OR) as suggested by the Basel Committee on Banking Supervision in 2003. In addition, recent E-banking financial losses together with risk management principles and standards raise the need for an effective ORA methodology and framework in the context of E-banking. Moreover, evaluation tools and / or methods for ORA are highly subjective, are still in their infant stages, and have not yet reached a consensus. Therefore, it is essential to develop valid and reliable methods for effective ORA and evaluations. The main contribution of this thesis is to apply Fuzzy Inference System (FIS) and Tree Augmented Naïve Bayes (TAN) classifier as standard tools for identifying OR, and measuring OR exposure level. In addition, a new ORA methodology is proposed which consists of four major steps: a risk model, assessment approach, analysis approach and a risk assessment process. Further, a new ORA framework and measurement metrics are proposed with six factors: frequency of triggering event, effectiveness of avoidance barriers, frequency of undesirable operational state, effectiveness of recovery barriers before the risk outcome, approximate cost for Undesirable Operational State (UOS) occurrence, and severity of the risk outcome. The study results were reported based on surveys conducted with Nigerian senior banking officers and banking customers. The study revealed that the framework and assessment tools gave good predictions for risk learning and inference in such systems. Thus, results obtained can be considered promising and useful for both E-banking system adopters and future researchers in this area.