In today‘s society, transport between places plays an important role in people‘s everyday life resulting in unavoidable effect on the environment. Emissions from the transport sector in the Nordic Region are expected to increase the coming years because of growing urbanization and population, so it becomes apparent that transport planners today are facing a complex system that requires to change in order to reach sustainability. Therefore, it is necessary to understand the planning systems and processes to enable improvements. The aim of this study is to analyse and compare how environmental aspects are incorporated in the process of defining measures at early stages in transport planning in the Nordic countries: Denmark, Finland, Iceland, Norway and Sweden, with the focus on each country‘s capital region. The focus is set on where and how collaboration between stakeholders and expert knowledge are included in the planning processes in the different countries. To meet the aim, a case study approach was chosen where semi-structured interviews with planning practitioners from all the countries were conducted to gather data, together with literature review and document analysis to set the scope of the study. The interview results showed that early-stage transport planning is practiced differently in the five countries. In Finland, Iceland, Norway and Sweden transport authorities or administrations are responsible for long-term strategic development of the transport system at a national and regional level, and they have defined how transport planning processes should be conducted. In these cases, collaboration between stakeholders and with experts occurs through workshops or working groups created at early stages. In Denmark, there is no longterm strategic transport planning at national or regional level. There, collaboration occurs at project level where experts and stakeholders are included when a project is being planned. Environmental assessment tools and methods were shown to be used at early stages and the interview results showed that CBA had a prominent role in Denmark and Norway, SEA and EIA had prominent roles in Iceland and an environmental assessment model in Finland. In Sweden, CBA and SEA are also used in transport planning, but in this study the focus was on Strategic Choice of Measures (SCM) which does not include those tools. In SCM, experts are included at early stages to incorporate environmental aspects and review chosen measures. Generally, the early-stage planning processes in the five countries are perceived as flexible which provides opportunities to adjust the processes to each case. Yet, planners need to motivate why a specific project or a solution is suggested and to do that they often use environmental assessment tools or expert knowledge.Integration of environmental concern in early planning stages has been identified as important in terms of reducing environmental problems in later stages. A suggestion for future studies is to investigate more in detail how different environmental assessment tools are used in the Nordic countries, as on what levels and stages they are applied. / Transporter spelar en viktig roll i människors vardag i dagens samhälle och bidrar till en oundviklig påverkan på miljön. Under de kommande åren förväntas utsläppen från transportsektorn i Norden öka till följd av urbanisering och stigande befolkningsmängd. Det bidrar till att transportplanerare idag står inför ett komplext system som kräver förändring för att kunna skapa en hållbar sektor. För att det ska ske är det viktigt att man förstår de bakomliggande planeringssystemen och processerna som kan bidra till förbättringar. Syftet med den här studien är att analysera och jämföra hur miljöaspekter integreras i ett tidigt skede där åtgärder definieras inom transportplaneringen i de nordiska länderna: Danmark, Finland, Island, Norge och Sverige, där fokus ligger på respektive lands huvudstadsregion. Fokus ligger även på var och hur samarbete sker mellan intressenter och experter i planeringsprocessen i de olika länderna. För att nå målen med studien har en fallstudie genomförts, där semistrukturerade intervjuermed planerare från samtliga länder har gjorts för att samla data, samt litteraturstudie och dokumentanalys för att bestämma studiens avgränsningar. Resultaten från intervjuerna visade att transportplaneringen i ett tidigt skede såg olika ut i de fem länderna. I Finland, Island, Norge och Sverige är transportmyndigheter eller styrelser ansvariga för att utveckla ett långsiktigt strategiskt transportsystem på national och regional nivå. De har även definierat hur transportplaneringsprocesser bör genomföras. I dessa fall sker samarbete mellan intressenter och experter genom seminarium eller arbetsgrupper skapade i ett tidigt skede. I Danmark finns ingen långsiktigt plan för strategiskt transportplanerande på nationell eller regional nivå. Samarbete mellan intressenter och experter sker istället på projektnivå. Olika verktyg och metoder för miljöbedömningar används i tidiga skeden, där kostnads-nyttoanalys används till största del i Danmark och Norge, strategisk miljöbedömning och miljökonsekvensbeskrivning används på Island och miljöbedömningsmodeller används i Finland. I Sverige genomförsocksåstrategisk miljöbedömningoch kostnads-nyttoanalys, meniden här studien är fokus på åtgärdsvalsstudier(ÅVS) i den svenska transportplaneringensom inte inkluderar de verktygen. I ÅVS är experter inkluderade från ett tidigt stadie för att inkludera miljöaspekter och utvärderaföreslagnaåtgärder. I de nordiska länderna uppfattas planering i ett tidigt skede under planeringsprocessen somflexibelvilket gör det möjligt att göra anpassningar till rådande planeringssituation. Trots flexibilitet behöverplanerare motiveravarför ett specifikt projekt eller lösningsförslag är framtagetoch för at göra det använder de ofta miljöbedömningsverktyg eller expertkunskap.Integrering av miljöhänsyn i tidiga planeringsstadier har identifierats som viktigt när det gäller att minska miljöproblem i senare skeden och för att åstadkomma ett hållbart transportsystem. Ett förslag för framtida studier är att undersöka mer i detalj hur olika miljöbedömningsverktyg används i de nordiska länderna, på vilka nivåer och stadier de används.

L'objectif initial de ce travail de thèse est de valoriser les informations relatives au suivi temporel de la vigne, du raisin et de l'environnement de la plante et enregistrées dans des bases de données (BD) de traçabilité pour permettre la comparaison entre parcelles et millésimes, en vue de décisions par analogies.Les travaux réalisés durant cette thèse ont permis de proposer une méthode de transformation qui permet de représenter des ensembles de données asynchrones dans un espace commun afin de les comparer. Cette méthode s'appuie sur l'expertise du système de production. Dans ce travail de thèse, cette méthode a été appliquée à la comparaison de couples parcelle×millésime.L'expertise du système de production viticole permet, dans une première phase, de définir (i) la forme générale de la cinétique d'évolution de grandeurs de mesures évaluées sur la vigne, le raisin ou l'environnement de la plante et (ii) l'effet du climat sur la plante. Cette expertise est utilisée, dans une seconde phase, pour proposer des modèles paramétriques de l'évolution de chaque grandeur. Les données de suivi de chaque couple parcelle×millésime permettent d'ajuster les paramètres du modèle. Un vecteur de paramètres est défini par couple parcelle×millésime. Ce vecteur représente l'espace commun qui rend les couples parcelle×millésime comparables. Deux stratégies de comparaison sontalors possibles : (i) les comparaisons sont réalisées à partir des paramètres (méthode intensive), ou (ii) les comparaisons sont réalisées à partir de l'estimation de la valeur de la grandeur pour chaque couple parcelle×millésime et chaque unité de temps, commune à tous les couples (méthode extensive).Cette méthode a été appliquée à trois exemples différents.Dans une première application, les climats de différents millésimes intervenus sur différents cépages, entre la floraison et la véraison, ont été comparés entre eux après modélisation des grandeurs de mesure climatiques, à l'aide de modèles très simples.Dans une seconde puis une troisième application, la cinétique d'augmentation du pH et d'accumulation des sucres dans les baies de raisin pendant la maturation a été modélisée sous la forme d'une sigmoïde. Les comparaisons ont ensuite été réalisées en travaillant sur (i) la courbe représentative de chaque cinétique (pH), (ii) les paramètres du modèle (sucres) et (iii) une estimation journalière de la concentration en sucres dans les baies.Les bases de données utilisées dans ces applications proviennent de deux régions très différentes. Des données issues du suivi de la station expérimentale INRA Pech Rouge en Languedoc-Roussillon, dans le sud de la France, ont été utilisées pour l'application 1 et une partie de l'application 3. Des données de suivi de différents domaines de la Napa Valley en Californie ont servi pour l'application 2 et une partie de l'application 3.Le changement d'espace de représentation des données apporte une connaissance nouvelle pour décrire les individus et les phénomènes temporels de la vigne. Cette connaissance pourrait permettre de formaliser un raisonnement par analogies utilisant et valorisant l'expérience passée pour la gestion du millésime en cours. / The initial objective of this thesis is to enhance the in-time follow-up information of the vine, the grape and the environment of the plant stored in traceability databases (BD) traceability to allow comparison between plots and vintages, to a future analogy-based decision support.The work done during this thesis allowed to propose a transformation method for representing sets of asynchronous data in a common space for comparison. This method relies on the expertise of the production system. In this thesis, this method was applied to the comparison of pairs of plot×vintage.The expert knowledge of the vineyard production system allows, in a first phase, to define (i) the general shape of the kinetics of on the vine, the grapes or the plant environment measured grandeurs, and (ii) the effect of the climate on the plant.This expertise is used in a second phase, to propose parametrical models that represent each grandeur kinetic. Monitoring data of each pair plot×vintage are used to adjust the model parameters. A vector of parameters is defined for each pair plot×vintage. This vector represents the common space that makes pairs of plot×vintage comparable. Two comparison strategies are possible: (i) comparisons are made from the parameters (intensive method), or (ii) comparisons are made from the estimation of the value of the quantity for each pair plot×vintage and each time unit, common to all pairs (extensive method).This method was applied to three different examples.In a first application, the climate of different vintages occurred on different grape varieties, between flowering and veraison, were compared with each other after modeling of the measured climate grandeurs, with very simple models.In a second and a third application, the kinetics of the increase of pH and accumulation of sugars in grape berries during ripening was modeled using a sigmoid. Comparisons were then made by working on (i) the graph of each kinetic (pH), (ii) parameters (sugars) and (iii) an estimation of the daily sugar concentration in berries.The databases used in these applications come from two very different winegrowing regions. Data from the monitoring of the INRA Pech Rouge Experimental Station, in Languedoc-Roussillon in the south of France, were used for the application 1 and part of the application 3. Monitoring data from different estates of Napa Valley in California were used for the application 2 and part of the application 3.The change of space where data are represented constitutes a new knowledge that permit one to describe individuals and temporal phenomena of the vine. This knowledge could allow to formalize an analogy-based reasoning that uses and promotes past experience to manage the current vintage.

This research aims to develop a process for eliciting expert knowledge and incorporating this knowledge as prior distributions for a Bayesian logistic regression model. This work was motivated by the need for less data reliant methods of modelling species habitat distributions. A comprehensive review of the research from both cognitive psychology and the statistical literature provided specific recommendations for the creation of an elicitation scheme. These were incorporated into the design of a Bayesian logistic regression model and accompanying elicitation scheme. This model and scheme were then implemented as interactive, graphical software called ELICITOR created within the BlackBox Component Pascal environment. This software was specifically written to be compatible with existing Bayesian analysis software, winBUGS as an odd-on component. The model, elicitation scheme and software were evaluated through five case studies of various fauna and flora species. For two of these there were sufficient data for a comparison of expert and data-driven models. The case studies confirmed that expert knowledge can be quantified and formally incorporated into a logistic regression model. Finally, they provide a basis for a thorough discussion of the model, scheme and software extensions and lead to recommendations for elicitation research.

This PhD thesis examines the status of ‘expert knowledge’ in the history of U.S. asymmetric, or ‘counterinsurgency’ (COIN), warfare during the last century. The historical rise of expert influence has so far been neglected in the study of wars within the field of International Relations and the thesis will give us an indication of the importance and utility of expert knowledge. With a specific focus on the campaigns in the Philippines (1899-1902), Vietnam (1954-75) and Iraq (2003-11), the central research question guiding the project is as follows: “What were the conditions for the evolution, the constitution and the use of ‘outside’ expert knowledge in U.S. counterinsurgency campaigns?” The thesis claims that military and academic ‘experts’ had a key role in framing and implementing the problem-sets and solutions to these conflicts. They have, in Iraq in particular, played an important part in developing the campaigns’ ex-post-facto justification of success. Within the framework of organisational knowledge production, this knowledge does not necessarily play an instrumental role for the military. Instead, it can also serve a merely symbolic function, demonstrating to the audience and stakeholders within the political environment that the organisation is willing to solve the problems the insurgents pose, but without any interest in long-term utilisation of the knowledge. This thesis argues that across time, from the beginning of the Philippine-American War in 1898 to the withdrawal of U.S. forces from Iraq in 2011, ‘counterinsurgency’ has developed from a tactical and operational tool, used instrumentally to fight insurgencies, to a strategy or even ‘ideology’ in its own right. Whilst the methods or techniques of counterinsurgency remain basically the same, expert knowledge is increasingly used in modern – that is post-World War II – campaigns to support a politico-strategic narrative.

Combine harvesters are used to gather plants from the field and separate them into the components of value, the grain and the straw. The optimal utilization of existing combine potential is an inevitable task to maximize harvest efficiency and hence to maximize profit. The only way to optimize the threshing and separation processes during harvest is to adjust the combine settings to existing conditions. Operating permanently at optimal harvest efficiency can only be achieved by an automatic control system. However, for reasons of transparency and due to lack of sensors, the approach in this thesis is a combined development of an interactive and an automatic control system for combine process optimization. The optimization of combine processes is a multi-dimensional and multi-objective optimization problem. The objectives of optimization are the harvest quality parameters. The decision variables, the parameters that can be modified, are the combine settings. Analytical optimization methods require the existence of a model that provides function values in dependence of defined input parameters. A comprehensive quantitative model for the input-output-behavior of the combine does not exist. Alternative optimization methods that handle multi-dimensional and multi-objective optimization problems can be found in the domain of Artificial Intelligence. In this work, knowledge acquisition was performed in order to obtain expert knowledge on combine process optimization. The result is a knowledge base with six adjustment matrices for different crop and combine types. The adjustment matrices contain problem oriented setting adjustment recommendations in order to solve single issues with quality parameters. A control algorithm has been developed that is also capable of solving multiple issues at the same time, utilizing the acquired expert knowledge. The basic principle to solve the given multi-objective optimization problem is a transformation into one-dimensional single-objective optimization problems which are solved iteratively. Several methods have been developed that are applied sequentially. In simulation, the average improvement from initial settings to optimized settings, achieved by the control algorithm, is between 34.5 % and 67.6 %. This demonstrates the good performance of the control algorithm.

This dissertation presents research on the value of using different sources of data to explore the factors determining invasiveness of introduced species. The research draws upon the availability of data on the historical trial plantings of alien species and other sources. The focus of the study is on Australian Acacia species as a taxon introduced into southern Africa (Lesotho, South Africa and Swaziland). The first component of the study focused on understanding the factors determining introduction outcome of species in historical trial plantings and invasion success of Australian Acacia species using Species Distribution Models (SDMs) and classification tree techniques. SDMs were calibrated using the native range occurrence records (Australia) and were validated using results of 150 years of South African government forestry trial planting records and invaded range data from the Southern African Plant Invaders Atlas. To understand factors associated with survival (‘trial success’) or failure to survive (‘trial failure’) of species in historical trial plantings, classification and regression tree analysis was used. The results indicate climate as one of the factors that explains introduction and/or invasion success of Australian Acacia species in southern Africa. However, the results also indicate that for ‘trial failures’ there are factors other than climate that could have influenced the trial outcome. This study emphasizes the need to integrate data on whether the species has been recorded to be invasive elsewhere with climate matching for invasion risk assessment. The second component of the study focused on understanding the distribution patterns of Australian Acacia species that are not known as invasive in southern Africa. The specific aims were to determine which species still exist at previously recorded sites and determine the current invasion status. This was done by collating data from different sources that list species introduced into southern Africa and then conducting revisits. For the purpose of this study, revisits means conducting field surveys based on recorded occurrences of introduced species. The known occurrence data for species on the list were obtained from different data sources and various invasion biology experts. As it was not practical to do revisits for all species on the list, three ornamental species (Acacia floribunda, A. pendula and A. retinodes) were selected as part of the pilot study for the conducted revisits in this study. Acacia retinodes trees were not found during the revisits. The results provided data that could be used to characterize species based on the Blackburn et al., (2011) scheme. However, it is not clear whether observed Acacia pendula or A. floribunda trees will spread away from the sites hence the need to continuously monitor sites for spread. The methods used in this research establish a protocol for future work on conducting revisits at known localities of introduced species to determine their population dynamics and thereby characterize the species according to the scheme for management purposes. / Dissertation (MSc)--University of Pretoria, 2014. / National Research Foundation (NRF) / Zoology and Entomology / MSc (Zoology) / Unrestricted

While machine learning has shown success in many fields, it can be challenging when there are limitations with insufficient training data. By incorporating knowledge into the machine learning pipeline, one can overcome such limitations. Therefore, eliciting expert knowledge can play an important role in the machine learning project pipeline. Expert knowledge can come in many forms, and it is seldom easy to elicit and formalize it in a way that is easily implementable into a machine learning project. While it has been done, not much focus has been on how. Furthermore, the motivations for why knowledge was elicited in a particular way as well as the challenges that may exist with the elicitation, are not always focused on either. Making educated decisions for knowledge elicitation can therefore be challenging for researchers. Hence, this work aims to explore and categorize how expert knowledge elicitation has been done by researchers previously. This was done by developing a taxonomy that was then used for analyzing articles. A total of 43 articles were found, containing 97 elicitation paths that were categorized in order to identify trends and common approaches. The findings from our study were used to provide guidance for an industrial case in its initial stage to show how the taxonomy presented in this work can be applied in a real-world scenario.

This dissertation explores the modular learning in artificial neural networks that mainly driven by the inspiration from the neurobiological basis of the human learning. The presented modularization approaches to the neural network design and learning are inspired by the engineering, complexity, psychological and neurobiological aspects. The main theme of this dissertation is to explore the organization and functioning of the brain to discover new structural and learning inspirations that can be subsequently utilized to design artificial neural network. The artificial neural networks are touted to be a neurobiologicaly inspired paradigm that emulate the functioning of the vertebrate brain. The brain is a highly structured entity with localized regions of neurons specialized in performing specific tasks. On the other hand, the mainstream monolithic feed-forward neural networks are generally unstructured black boxes which is their major performance limiting characteristic. The non explicit structure and monolithic nature of the current mainstream artificial neural networks results in lack of the capability of systematic incorporation of functional or task-specific a priori knowledge in the artificial neural network design process. The problem caused by these limitations are discussed in detail in this dissertation and remedial solutions are presented that are driven by the functioning of the brain and its structural organization. Also, this dissertation presents an in depth study of the currently available modular neural network architectures along with highlighting their shortcomings and investigates new modular artificial neural network models in order to overcome pointed out shortcomings. The resulting proposed modular neural network models have greater accuracy, generalization, comprehensible simplified neural structure, ease of training and more user confidence. These benefits are readily obvious for certain problems, depending upon availability and usage of available a priori knowledge about the problems. The modular neural network models presented in this dissertation exploit the capabilities of the principle of divide and conquer in the design and learning of the modular artificial neural networks. The strategy of divide and conquer solves a complex computational problem by dividing it into simpler sub-problems and then combining the individual solutions to the sub-problems into a solution to the original problem. The divisions of a task considered in this dissertation are the automatic decomposition of the mappings to be learned, decompositions of the artificial neural networks to minimize harmful interaction during the learning process, and explicit decomposition of the application task into sub-tasks that are learned separately. The versatility and capabilities of the new proposed modular neural networks are demonstrated by the experimental results. A comparison of the current modular neural network design techniques with the ones introduced in this dissertation, is also presented for reference. The results presented in this dissertation lay a solid foundation for design and learning of the artificial neural networks that have sound neurobiological basis that leads to superior design techniques. Areas of the future research are also presented. / Ph. D.

This thesis deals with the integration of auxiliary data knowledge into machine learning methods especially prototype based classification models. The problem of classification is diverse and evaluation of the result by using only the accuracy is not adequate in many applications. Therefore, the classification tasks are analyzed more deeply. Possibilities to extend prototype based methods to integrate extra knowledge about the data or the classification goal is presented to obtain problem adequate models. One of the proposed extensions is Generalized Learning Vector Quantization for direct optimization of statistical measurements besides the classification accuracy. But also modifying the metric adaptation of the Generalized Learning Vector Quantization for functional data, i. e. data with lateral dependencies in the features, is considered.

This thesis focuses on the problem of scalable optimization of dialogue behaviour in speech-based conversational systems using reinforcement learning. Most previous investigations in dialogue strategy learning have proposed flat reinforcement learning methods, which are more suitable for small-scale spoken dialogue systems. This research formulates the problem in terms of Semi-Markov Decision Processes (SMDPs), and proposes two hierarchical reinforcement learning methods to optimize sub-dialogues rather than full dialogues. The first method uses a hierarchy of SMDPs, where every SMDP ignores irrelevant state variables and actions in order to optimize a sub-dialogue. The second method extends the first one by constraining every SMDP in the hierarchy with prior expert knowledge. The latter method proposes a learning algorithm called 'HAM+HSMQ-Learning', which combines two existing algorithms in the literature of hierarchical reinforcement learning. Whilst the first method generates fully-learnt behaviour, the second one generates semi-learnt behaviour. In addition, this research proposes a heuristic dialogue simulation environment for automatic dialogue strategy learning. Experiments were performed on simulated and real environments based on a travel planning spoken dialogue system. Experimental results provided evidence to support the following claims: First, both methods scale well at the cost of near-optimal solutions, resulting in slightly longer dialogues than the optimal solutions. Second, dialogue strategies learnt with coherent user behaviour and conservative recognition error rates can outperform a reasonable hand-coded strategy. Third, semi-learnt dialogue behaviours are a better alternative (because of their higher overall performance) than hand-coded or fully-learnt dialogue behaviours. Last, hierarchical reinforcement learning dialogue agents are feasible and promising for the (semi) automatic design of adaptive behaviours in larger-scale spoken dialogue systems. This research makes the following contributions to spoken dialogue systems which learn their dialogue behaviour. First, the Semi-Markov Decision Process (SMDP) model was proposed to learn spoken dialogue strategies in a scalable way. Second, the concept of 'partially specified dialogue strategies' was proposed for integrating simultaneously hand-coded and learnt spoken dialogue behaviours into a single learning framework. Third, an evaluation with real users of hierarchical reinforcement learning dialogue agents was essential to validate their effectiveness in a realistic environment.