Could machine learning improve the prediction of pelvic nodal status of prostate cancer patients? Preliminary results of a pilot study.

We tested and compared performances of Roach formula, Partin tables and of three Machine Learning (ML) based algorithms based on decision trees in identifying N+ prostate cancer (PC). 1,555 cN0 and 50 cN+ PC were analyzed. Results were also verified on an independent population of 204 operated cN0 patients, with a known pN status (187 pN0, 17 pN1 patients). ML performed better, also when tested on the surgical population, with accuracy, specificity, and sensitivity ranging between 48-86%, 35-91%, and 17-79%, respectively. ML potentially allows better prediction of the nodal status of PC, potentially allowing a better tailoring of pelvic irradiation.

Error models in hydrogeology applications

Notre consommation en eau souterraine, en particulier comme eau potable ou pour l'irrigation, a considérablement augmenté au cours des années. De nombreux problèmes font alors leur apparition, allant de la prospection de nouvelles ressources à la remédiation des aquifères pollués. Indépendamment du problème hydrogéologique considéré, le principal défi reste la caractérisation des propriétés du sous-sol. Une approche stochastique est alors nécessaire afin de représenter cette incertitude en considérant de multiples scénarios géologiques et en générant un grand nombre de réalisations géostatistiques. Nous rencontrons alors la principale limitation de ces approches qui est le coût de calcul dû à la simulation des processus d'écoulements complexes pour chacune de ces réalisations. Dans la première partie de la thèse, ce problème est investigué dans le contexte de propagation de l'incertitude, oú un ensemble de réalisations est identifié comme représentant les propriétés du sous-sol. Afin de propager cette incertitude à la quantité d'intérêt tout en limitant le coût de calcul, les méthodes actuelles font appel à des modèles d'écoulement approximés. Cela permet l'identification d'un sous-ensemble de réalisations représentant la variabilité de l'ensemble initial. Le modèle complexe d'écoulement est alors évalué uniquement pour ce sousensemble, et, sur la base de ces réponses complexes, l'inférence est faite. Notre objectif est d'améliorer la performance de cette approche en utilisant toute l'information à disposition. Pour cela, le sous-ensemble de réponses approximées et exactes est utilisé afin de construire un modèle d'erreur, qui sert ensuite à corriger le reste des réponses approximées et prédire la réponse du modèle complexe. Cette méthode permet de maximiser l'utilisation de l'information à disposition sans augmentation perceptible du temps de calcul. La propagation de l'incertitude est alors plus précise et plus robuste. La stratégie explorée dans le premier chapitre consiste à apprendre d'un sous-ensemble de réalisations la relation entre les modèles d'écoulement approximé et complexe. Dans la seconde partie de la thèse, cette méthodologie est formalisée mathématiquement en introduisant un modèle de régression entre les réponses fonctionnelles. Comme ce problème est mal posé, il est nécessaire d'en réduire la dimensionnalité. Dans cette optique, l'innovation du travail présenté provient de l'utilisation de l'analyse en composantes principales fonctionnelles (ACPF), qui non seulement effectue la réduction de dimensionnalités tout en maximisant l'information retenue, mais permet aussi de diagnostiquer la qualité du modèle d'erreur dans cet espace fonctionnel. La méthodologie proposée est appliquée à un problème de pollution par une phase liquide nonaqueuse et les résultats obtenus montrent que le modèle d'erreur permet une forte réduction du temps de calcul tout en estimant correctement l'incertitude. De plus, pour chaque réponse approximée, une prédiction de la réponse complexe est fournie par le modèle d'erreur. Le concept de modèle d'erreur fonctionnel est donc pertinent pour la propagation de l'incertitude, mais aussi pour les problèmes d'inférence bayésienne. Les méthodes de Monte Carlo par chaîne de Markov (MCMC) sont les algorithmes les plus communément utilisés afin de générer des réalisations géostatistiques en accord avec les observations. Cependant, ces méthodes souffrent d'un taux d'acceptation très bas pour les problèmes de grande dimensionnalité, résultant en un grand nombre de simulations d'écoulement gaspillées. Une approche en deux temps, le "MCMC en deux étapes", a été introduite afin d'éviter les simulations du modèle complexe inutiles par une évaluation préliminaire de la réalisation. Dans la troisième partie de la thèse, le modèle d'écoulement approximé couplé à un modèle d'erreur sert d'évaluation préliminaire pour le "MCMC en deux étapes". Nous démontrons une augmentation du taux d'acceptation par un facteur de 1.5 à 3 en comparaison avec une implémentation classique de MCMC. Une question reste sans réponse : comment choisir la taille de l'ensemble d'entrainement et comment identifier les réalisations permettant d'optimiser la construction du modèle d'erreur. Cela requiert une stratégie itérative afin que, à chaque nouvelle simulation d'écoulement, le modèle d'erreur soit amélioré en incorporant les nouvelles informations. Ceci est développé dans la quatrième partie de la thèse, oú cette méthodologie est appliquée à un problème d'intrusion saline dans un aquifère côtier. -- Our consumption of groundwater, in particular as drinking water and for irrigation, has considerably increased over the years and groundwater is becoming an increasingly scarce and endangered resource. Nofadays, we are facing many problems ranging from water prospection to sustainable management and remediation of polluted aquifers. Independently of the hydrogeological problem, the main challenge remains dealing with the incomplete knofledge of the underground properties. Stochastic approaches have been developed to represent this uncertainty by considering multiple geological scenarios and generating a large number of realizations. The main limitation of this approach is the computational cost associated with performing complex of simulations in each realization. In the first part of the thesis, we explore this issue in the context of uncertainty propagation, where an ensemble of geostatistical realizations is identified as representative of the subsurface uncertainty. To propagate this lack of knofledge to the quantity of interest (e.g., the concentration of pollutant in extracted water), it is necessary to evaluate the of response of each realization. Due to computational constraints, state-of-the-art methods make use of approximate of simulation, to identify a subset of realizations that represents the variability of the ensemble. The complex and computationally heavy of model is then run for this subset based on which inference is made. Our objective is to increase the performance of this approach by using all of the available information and not solely the subset of exact responses. Two error models are proposed to correct the approximate responses follofing a machine learning approach. For the subset identified by a classical approach (here the distance kernel method) both the approximate and the exact responses are knofn. This information is used to construct an error model and correct the ensemble of approximate responses to predict the "expected" responses of the exact model. The proposed methodology makes use of all the available information without perceptible additional computational costs and leads to an increase in accuracy and robustness of the uncertainty propagation. The strategy explored in the first chapter consists in learning from a subset of realizations the relationship between proxy and exact curves. In the second part of this thesis, the strategy is formalized in a rigorous mathematical framework by defining a regression model between functions. As this problem is ill-posed, it is necessary to reduce its dimensionality. The novelty of the work comes from the use of functional principal component analysis (FPCA), which not only performs the dimensionality reduction while maximizing the retained information, but also allofs a diagnostic of the quality of the error model in the functional space. The proposed methodology is applied to a pollution problem by a non-aqueous phase-liquid. The error model allofs a strong reduction of the computational cost while providing a good estimate of the uncertainty. The individual correction of the proxy response by the error model leads to an excellent prediction of the exact response, opening the door to many applications. The concept of functional error model is useful not only in the context of uncertainty propagation, but also, and maybe even more so, to perform Bayesian inference. Monte Carlo Markov Chain (MCMC) algorithms are the most common choice to ensure that the generated realizations are sampled in accordance with the observations. Hofever, this approach suffers from lof acceptance rate in high dimensional problems, resulting in a large number of wasted of simulations. This led to the introduction of two-stage MCMC, where the computational cost is decreased by avoiding unnecessary simulation of the exact of thanks to a preliminary evaluation of the proposal. In the third part of the thesis, a proxy is coupled to an error model to provide an approximate response for the two-stage MCMC set-up. We demonstrate an increase in acceptance rate by a factor three with respect to one-stage MCMC results. An open question remains: hof do we choose the size of the learning set and identify the realizations to optimize the construction of the error model. This requires devising an iterative strategy to construct the error model, such that, as new of simulations are performed, the error model is iteratively improved by incorporating the new information. This is discussed in the fourth part of the thesis, in which we apply this methodology to a problem of saline intrusion in a coastal aquifer.

Chest CT at a dose below 0.3 mSv : impact of iterative reconstruction on image quality and lung analysis

Le nombre d'examens tomodensitométriques (Computed Tomography, CT) effectués chaque année étant en constante augmentation, différentes techniques d'optimisation, dont les algorithmes de reconstruction itérative permettant de réduire le bruit tout en maintenant la résolution spatiale, ont étés développées afin de réduire les doses délivrées. Le but de cette étude était d'évaluer l'impact des algorithmes de reconstruction itérative sur la qualité image à des doses effectives inférieures à 0.3 mSv, comparables à celle d'une radiographie thoracique. Vingt CT thoraciques effectués à cette dose effective ont été reconstruits en variant trois paramètres: l'algorithme de reconstruction, rétroprojection filtrée versus reconstruction itérative iDose4; la matrice, 5122 versus 7682; et le filtre de résolution en densité (mou) versus spatiale (dur). Ainsi, 8 séries ont été reconstruites pour chacun des 20 CT thoraciques. La qualité d'image de ces 8 séries a d'abord été évaluée qualitativement par deux radiologues expérimentés en aveugle en se basant sur la netteté des parois bronchiques et de l'interface entre le parenchyme pulmonaire et les vaisseaux, puis quantitativement en utilisant une formule de merit, fréquemment utilisée dans le développement de nouveaux algorithmes et filtres de reconstruction. La performance diagnostique de la meilleure série acquise à une dose effective inférieure à 0.3 mSv a été comparée à celle d'un CT de référence effectué à doses standards en relevant les anomalies du parenchyme pulmonaire. Les résultats montrent que la meilleure qualité d'image, tant qualitativement que quantitativement a été obtenue en utilisant iDose4, la matrice 5122 et le filtre mou, avec une concordance parfaite entre les classements quantitatif et qualitatif des 8 séries. D'autre part, la détection des nodules pulmonaires de plus de 4mm étaient similaire sur la meilleure série acquise à une dose effective inférieure à 0.3 mSv et le CT de référence. En conclusion, les CT thoraciques effectués à une dose effective inférieure à 0.3 mSv reconstruits avec iDose4, la matrice 5122 et le filtre mou peuvent être utilisés avec confiance pour diagnostiquer les nodules pulmonaires de plus de 4mm.

Knowledge creation and transfer in an iterative project environment

This Master’s Thesis examines knowledge creation and transfer processes in an iterative project environment. The aim is to understand how knowledge is created and transferred during an actual iterative implementation project which takes place in International Business Machines (IBM). The second aim is to create and develop new working methods that support more effective knowledge creation and transfer for future iterative implementation projects. The research methodology in this thesis is qualitative. Using focus group interviews as a research method provides qualitative information and introduces the experiences of the individuals participating in the project. This study found that the following factors affect knowledge creation and transfer in an iterative, multinational, and multi-organizational implementation project: shared vision and common goal, trust, open communication, social capital, and network density. All of these received both theoretical and empirical support. As for future projects, strengthening these factors was found to be the key for more effective knowledge creation and transfer.

Los enfoques de aprendizaje en estudiantes universitarios Catalanes mediante el approaches and study skills inventory for students (ASSIST) = Learning Approaches of Catalan University Students Measured with the Approaches and Study Skills Inventory for Students (ASSIST)

The aim of this study is to measure the psychometric properties of a Catalan translation of the Approaches and Study Skills Inventory for Students (ASSIST), and to analyse the different learning styles used by university students, considering the influence of gender and type of studies. The instrument was administered to 834 students at the University of Girona. The results showed that most students interviewed had a deep approach to learning, although the analysis by gender showed that females tended to use a more strategic approach, while males used a deep approach predominantly. As to whether the type of studies influenced learning styles, a prevalence of deep approach was found among Science and Technology students, while a more strategic approach was found among Humanities and Education students

New kernel functions and learning methods for text and data mining

Recent advances in machine learning methods enable increasingly the automatic construction of various types of computer assisted methods that have been difficult or laborious to program by human experts. The tasks for which this kind of tools are needed arise in many areas, here especially in the fields of bioinformatics and natural language processing. The machine learning methods may not work satisfactorily if they are not appropriately tailored to the task in question. However, their learning performance can often be improved by taking advantage of deeper insight of the application domain or the learning problem at hand. This thesis considers developing kernel-based learning algorithms incorporating this kind of prior knowledge of the task in question in an advantageous way. Moreover, computationally efficient algorithms for training the learning machines for specific tasks are presented. In the context of kernel-based learning methods, the incorporation of prior knowledge is often done by designing appropriate kernel functions. Another well-known way is to develop cost functions that fit to the task under consideration. For disambiguation tasks in natural language, we develop kernel functions that take account of the positional information and the mutual similarities of words. It is shown that the use of this information significantly improves the disambiguation performance of the learning machine. Further, we design a new cost function that is better suitable for the task of information retrieval and for more general ranking problems than the cost functions designed for regression and classification. We also consider other applications of the kernel-based learning algorithms such as text categorization, and pattern recognition in differential display. We develop computationally efficient algorithms for training the considered learning machines with the proposed kernel functions. We also design a fast cross-validation algorithm for regularized least-squares type of learning algorithm. Further, an efficient version of the regularized least-squares algorithm that can be used together with the new cost function for preference learning and ranking tasks is proposed. In summary, we demonstrate that the incorporation of prior knowledge is possible and beneficial, and novel advanced kernels and cost functions can be used in algorithms efficiently.

Análisis de una experiencia de aprendizaje basado en proyectos en la asignatura de Fundamentos de Física = Project based learning: analysis of an experience in Introductory Physics courses

In this work we present and analyze the application of an experience of Project Based Learning (PBL) in the matter of Physics II of the Industrial Design university degree (Girona University) during 2005-2006 courses. This methodology was applied to the Electrostatic and Direct Current subjects. Furthermore, evaluation and self evaluation results were shown and the academic results were compared with results obtained in the same subjects applying conventional teaching methods

Gestión de la actividad no presencial y autoevaluación de competencias = Management of Student Learning Outside the Classroom and the Self-evaluation of Competences

Psychological Assessment is a core subject of Psychology studies, and of the university degree Human development, at the University of Girona and according to the University regulations, had 12 credits. Until the 2005-06 academic years, the student work conducted outside the classroom consisted of undertaking a psychological assessment that was written and handed in at the end of the course. From this the student obtained a qualification and a review if they applied for it In accordance with the European Credits for Higher Education, this subject was reduced to 9 credits, which is the equivalent of a total of 255 hours of in-class work and outside the classroom. In the 2006-07 academic year we created a guide to manage the student’s experiences outside the classroom, with the objective of encouraging the application of problem solving/critical thinking (Bloom, 1975), in line with the recommendations of the Catalonia Agency for University System Quality (2005). The guide includes: learning objectives, evaluation criterions, a description of activities, work week timetable for the whole course, programmed tutorials to review all steps of the psychological assessment process, and the use of a web-based virtual forum for the transfer of knowledge, analysis and constructive critiques of the assessment done by themselves and their colleagues

The motivation to learn as a self-presentation tool among Swiss high school students : The moderating role of mastery goals' perceived social value on learning

Although it has been assumed that the motivation to learn - or mastery goal endorsement - positively predicts learning achievement, most empirical findings fail to demonstrate this relationship. In the present research, conducted in a Swiss high school, we adopted a social value approach to test the hypothesis that adolescent students' mastery goals do in fact predict learning, but only if these goals are perceived as highly useful for scholarly success (high social utility), and are not endorsed as a means to be appreciated by the teachers (low social desirability), a finding that has previously been observed among college students and on teacher-graded achievement measures only. Results demonstrate that in spite of potential peculiarities of an adolescent population, individual differences in mastery goals' perceived social utility and desirability moderate the mastery goal endorsement-learning achievement relation. Findings are discussed with regard to both theory development and educational practice.

Literature report : a review of policy learning in five strands of political science research

This paper reviews the policy learning literature in political science. In recent years, the number of publications on learning in the political realm increased dramatically. Researchers have focused on how policymakers and administrators adapt policies based on learning processes or experiences. Thereby, learning has been discussed in very different ways. Authors have referred to learning in the context of ideas, understood as deeply held beliefs, and, as change and adaptation of policy instruments. Two other strands of literature have taken into consideration learning, namely the diffusion literature and research on transfer, which put forward learning to understand who learns from whom and what. Opposed to these views, political learning emphasizes the adaptation of new strategies by policymakers over the transfer of knowledge or broad ideas. In this approach, learning occurs due to the failure of existing policies, increasing problem pressure, scientific innovations, or a combination of these elements.

What can ecosystems learn? Expanding evolutionary ecology with learning theory.

BACKGROUND: The structure and organisation of ecological interactions within an ecosystem is modified by the evolution and coevolution of the individual species it contains. Understanding how historical conditions have shaped this architecture is vital for understanding system responses to change at scales from the microbial upwards. However, in the absence of a group selection process, the collective behaviours and ecosystem functions exhibited by the whole community cannot be organised or adapted in a Darwinian sense. A long-standing open question thus persists: Are there alternative organising principles that enable us to understand and predict how the coevolution of the component species creates and maintains complex collective behaviours exhibited by the ecosystem as a whole? RESULTS: Here we answer this question by incorporating principles from connectionist learning, a previously unrelated discipline already using well-developed theories on how emergent behaviours arise in simple networks. Specifically, we show conditions where natural selection on ecological interactions is functionally equivalent to a simple type of connectionist learning, 'unsupervised learning', well-known in neural-network models of cognitive systems to produce many non-trivial collective behaviours. Accordingly, we find that a community can self-organise in a well-defined and non-trivial sense without selection at the community level; its organisation can be conditioned by past experience in the same sense as connectionist learning models habituate to stimuli. This conditioning drives the community to form a distributed ecological memory of multiple past states, causing the community to: a) converge to these states from any random initial composition; b) accurately restore historical compositions from small fragments; c) recover a state composition following disturbance; and d) to correctly classify ambiguous initial compositions according to their similarity to learned compositions. We examine how the formation of alternative stable states alters the community's response to changing environmental forcing, and we identify conditions under which the ecosystem exhibits hysteresis with potential for catastrophic regime shifts. CONCLUSIONS: This work highlights the potential of connectionist theory to expand our understanding of evo-eco dynamics and collective ecological behaviours. Within this framework we find that, despite not being a Darwinian unit, ecological communities can behave like connectionist learning systems, creating internal conditions that habituate to past environmental conditions and actively recalling those conditions. REVIEWERS: This article was reviewed by Prof. Ricard V Solé, Universitat Pompeu Fabra, Barcelona and Prof. Rob Knight, University of Colorado, Boulder.

Labour market innovations and policy learning : report on Switzerland

This paper analyses learning and implementation of labour market reforms in Switzerland.

Best practice report on policy learning infrastructures in innovative labour market policies

This report synthesizes the findings of 11 country reports on policy learning in labour market and social policies that were conducted as part of WP5 of the INSPIRES project, which is funded by the 7th Framework Program of the EU-Commission. Notably, this report puts forward objectives of policy learning, discusses tools, processes and institutions of policy learning and presents the impacts of various tools and structures of the policy learning infrastructure for the actual policy learning process. The report defines three objectives of policy learning: evaluation and assessment of policy effectiveness, vision building and planning, and consensus building. In the 11 countries under consideration, the tools and processes of the policy learning, infrastructure can be classified into three broad groups: public bodies, expert councils, and parties, interest groups and the private sector. Finally, we develop four recommendations for policy learning: Firstly, learning processes should keep the balance between centralisation and plurality. Secondly, learning processes should be kept stable beyond the usual political business cycles. Thirdly, policy learning tools and infrastructures should be sufficiently independent from political influence or bias. Fourth, Policy learning tools and infrastructures should balance out mere effectiveness, evaluation and vision building.

Policy learning in Europe : social policy and labour market reforms in 11 countries

This report compares policy learning processes in 11 European countries. Based on the country reports that were produced by the national teams of the INSPIRES project, this paper develops an argument that connects problem pressure and politicization to learning in different labor market innovations. In short, we argue that learning efforts are most likely to impact on policy change if there is a certain problem pressure that clearly necessitates political action. On the other hand, if problem pressure is very low, or so high that governments need to react immediately, chances are low that learning impacts on policy change. The second part of our argument contends that learning impacts on policy change especially if a problem is not very politicized, i.e. there are no main conflicts concerning a reform, because then, solutions are wound up in the search for a compromise. Our results confirm our first hypothesis regarding the connection between problem pressure and policy learning. Governments learn indeed up to a certain degree of problem pressure. However, once political action becomes really urgent, i.e. in anti-crisis policies, there is no time and room for learning. On the other hand, learning occurred independently from the politicization of problem. In fact, in countries that have a consensual political system, learning occurred before the decision on a reform, whereas in majoritarian systems, learning happened after the adoption of a policy during the process of implementation.