Machine learning for geospatial data : algorithms, software tools and case studies

Résumé Cette thèse est consacrée à l'analyse, la modélisation et la visualisation de données environnementales à référence spatiale à l'aide d'algorithmes d'apprentissage automatique (Machine Learning). L'apprentissage automatique peut être considéré au sens large comme une sous-catégorie de l'intelligence artificielle qui concerne particulièrement le développement de techniques et d'algorithmes permettant à une machine d'apprendre à partir de données. Dans cette thèse, les algorithmes d'apprentissage automatique sont adaptés pour être appliqués à des données environnementales et à la prédiction spatiale. Pourquoi l'apprentissage automatique ? Parce que la majorité des algorithmes d'apprentissage automatiques sont universels, adaptatifs, non-linéaires, robustes et efficaces pour la modélisation. Ils peuvent résoudre des problèmes de classification, de régression et de modélisation de densité de probabilités dans des espaces à haute dimension, composés de variables informatives spatialisées (« géo-features ») en plus des coordonnées géographiques. De plus, ils sont idéaux pour être implémentés en tant qu'outils d'aide à la décision pour des questions environnementales allant de la reconnaissance de pattern à la modélisation et la prédiction en passant par la cartographie automatique. Leur efficacité est comparable au modèles géostatistiques dans l'espace des coordonnées géographiques, mais ils sont indispensables pour des données à hautes dimensions incluant des géo-features. Les algorithmes d'apprentissage automatique les plus importants et les plus populaires sont présentés théoriquement et implémentés sous forme de logiciels pour les sciences environnementales. Les principaux algorithmes décrits sont le Perceptron multicouches (MultiLayer Perceptron, MLP) - l'algorithme le plus connu dans l'intelligence artificielle, le réseau de neurones de régression généralisée (General Regression Neural Networks, GRNN), le réseau de neurones probabiliste (Probabilistic Neural Networks, PNN), les cartes auto-organisées (SelfOrganized Maps, SOM), les modèles à mixture Gaussiennes (Gaussian Mixture Models, GMM), les réseaux à fonctions de base radiales (Radial Basis Functions Networks, RBF) et les réseaux à mixture de densité (Mixture Density Networks, MDN). Cette gamme d'algorithmes permet de couvrir des tâches variées telle que la classification, la régression ou l'estimation de densité de probabilité. L'analyse exploratoire des données (Exploratory Data Analysis, EDA) est le premier pas de toute analyse de données. Dans cette thèse les concepts d'analyse exploratoire de données spatiales (Exploratory Spatial Data Analysis, ESDA) sont traités selon l'approche traditionnelle de la géostatistique avec la variographie expérimentale et selon les principes de l'apprentissage automatique. La variographie expérimentale, qui étudie les relations entre pairs de points, est un outil de base pour l'analyse géostatistique de corrélations spatiales anisotropiques qui permet de détecter la présence de patterns spatiaux descriptible par une statistique. L'approche de l'apprentissage automatique pour l'ESDA est présentée à travers l'application de la méthode des k plus proches voisins qui est très simple et possède d'excellentes qualités d'interprétation et de visualisation. Une part importante de la thèse traite de sujets d'actualité comme la cartographie automatique de données spatiales. Le réseau de neurones de régression généralisée est proposé pour résoudre cette tâche efficacement. Les performances du GRNN sont démontrées par des données de Comparaison d'Interpolation Spatiale (SIC) de 2004 pour lesquelles le GRNN bat significativement toutes les autres méthodes, particulièrement lors de situations d'urgence. La thèse est composée de quatre chapitres : théorie, applications, outils logiciels et des exemples guidés. Une partie importante du travail consiste en une collection de logiciels : Machine Learning Office. Cette collection de logiciels a été développée durant les 15 dernières années et a été utilisée pour l'enseignement de nombreux cours, dont des workshops internationaux en Chine, France, Italie, Irlande et Suisse ainsi que dans des projets de recherche fondamentaux et appliqués. Les cas d'études considérés couvrent un vaste spectre de problèmes géoenvironnementaux réels à basse et haute dimensionnalité, tels que la pollution de l'air, du sol et de l'eau par des produits radioactifs et des métaux lourds, la classification de types de sols et d'unités hydrogéologiques, la cartographie des incertitudes pour l'aide à la décision et l'estimation de risques naturels (glissements de terrain, avalanches). Des outils complémentaires pour l'analyse exploratoire des données et la visualisation ont également été développés en prenant soin de créer une interface conviviale et facile à l'utilisation. Machine Learning for geospatial data: algorithms, software tools and case studies Abstract The thesis is devoted to the analysis, modeling and visualisation of spatial environmental data using machine learning algorithms. In a broad sense machine learning can be considered as a subfield of artificial intelligence. It mainly concerns with the development of techniques and algorithms that allow computers to learn from data. In this thesis machine learning algorithms are adapted to learn from spatial environmental data and to make spatial predictions. Why machine learning? In few words most of machine learning algorithms are universal, adaptive, nonlinear, robust and efficient modeling tools. They can find solutions for the classification, regression, and probability density modeling problems in high-dimensional geo-feature spaces, composed of geographical space and additional relevant spatially referenced features. They are well-suited to be implemented as predictive engines in decision support systems, for the purposes of environmental data mining including pattern recognition, modeling and predictions as well as automatic data mapping. They have competitive efficiency to the geostatistical models in low dimensional geographical spaces but are indispensable in high-dimensional geo-feature spaces. The most important and popular machine learning algorithms and models interesting for geo- and environmental sciences are presented in details: from theoretical description of the concepts to the software implementation. The main algorithms and models considered are the following: multi-layer perceptron (a workhorse of machine learning), general regression neural networks, probabilistic neural networks, self-organising (Kohonen) maps, Gaussian mixture models, radial basis functions networks, mixture density networks. This set of models covers machine learning tasks such as classification, regression, and density estimation. Exploratory data analysis (EDA) is initial and very important part of data analysis. In this thesis the concepts of exploratory spatial data analysis (ESDA) is considered using both traditional geostatistical approach such as_experimental variography and machine learning. Experimental variography is a basic tool for geostatistical analysis of anisotropic spatial correlations which helps to understand the presence of spatial patterns, at least described by two-point statistics. A machine learning approach for ESDA is presented by applying the k-nearest neighbors (k-NN) method which is simple and has very good interpretation and visualization properties. Important part of the thesis deals with a hot topic of nowadays, namely, an automatic mapping of geospatial data. General regression neural networks (GRNN) is proposed as efficient model to solve this task. Performance of the GRNN model is demonstrated on Spatial Interpolation Comparison (SIC) 2004 data where GRNN model significantly outperformed all other approaches, especially in case of emergency conditions. The thesis consists of four chapters and has the following structure: theory, applications, software tools, and how-to-do-it examples. An important part of the work is a collection of software tools - Machine Learning Office. Machine Learning Office tools were developed during last 15 years and was used both for many teaching courses, including international workshops in China, France, Italy, Ireland, Switzerland and for realizing fundamental and applied research projects. Case studies considered cover wide spectrum of the real-life low and high-dimensional geo- and environmental problems, such as air, soil and water pollution by radionuclides and heavy metals, soil types and hydro-geological units classification, decision-oriented mapping with uncertainties, natural hazards (landslides, avalanches) assessments and susceptibility mapping. Complementary tools useful for the exploratory data analysis and visualisation were developed as well. The software is user friendly and easy to use.

Diseño e implementación de un framework de presentación : El framework Myst

Estudio sobre los frameworks Java de presentación. Creación de un framework de presentación propio y de una aplicación JEE que lo utiliza.

Disseny i implementació d'un framework de persistència

Aquest projecte consisteix en la realització d'un framework de persistència per realitzar l'enllaç entre el model de dades relacional i el model de dades orientat a objectes. L'aplicació que implementarem per tal de provar el nostre framework consisteix en l'adaptació web d'un software financer que serveix per tenir un control de la tresoreria.

Disseny i implementació d'un framework de persistència

En aquest PFC s'analitzaran els diversos frameworks de persistència que actualment existeixen i es facilitarà el desenvolupament d'un conjunt de components que permetran simplificar la capa de persistència en aplicacions multicapa desenvolupades amb JEE.

A Framework for Building Hypercubes Using MapReduce

The European Space Agency's Gaia mission will create the largest and most precise three dimensional chart of our galaxy (the Milky Way), by providing unprecedented position, parallax, proper motion, and radial velocity measurements for about one billion stars. The resulting catalogue will be made available to the scientific community and will be analyzed in many different ways, including the production of a variety of statistics. The latter will often entail the generation of multidimensional histograms and hypercubes as part of the precomputed statistics for each data release, or for scientific analysis involving either the final data products or the raw data coming from the satellite instruments. In this paper we present and analyze a generic framework that allows the hypercube generation to be easily done within a MapReduce infrastructure, providing all the advantages of the new Big Data analysis paradigmbut without dealing with any specific interface to the lower level distributed system implementation (Hadoop). Furthermore, we show how executing the framework for different data storage model configurations (i.e. row or column oriented) and compression techniques can considerably improve the response time of this type of workload for the currently available simulated data of the mission. In addition, we put forward the advantages and shortcomings of the deployment of the framework on a public cloud provider, benchmark against other popular solutions available (that are not always the best for such ad-hoc applications), and describe some user experiences with the framework, which was employed for a number of dedicated astronomical data analysis techniques workshops.

A Framework for Building Hypercubes Using MapReduce

The European Space Agency's Gaia mission will create the largest and most precise three dimensional chart of our galaxy (the Milky Way), by providing unprecedented position, parallax, proper motion, and radial velocity measurements for about one billion stars. The resulting catalogue will be made available to the scientific community and will be analyzed in many different ways, including the production of a variety of statistics. The latter will often entail the generation of multidimensional histograms and hypercubes as part of the precomputed statistics for each data release, or for scientific analysis involving either the final data products or the raw data coming from the satellite instruments. In this paper we present and analyze a generic framework that allows the hypercube generation to be easily done within a MapReduce infrastructure, providing all the advantages of the new Big Data analysis paradigmbut without dealing with any specific interface to the lower level distributed system implementation (Hadoop). Furthermore, we show how executing the framework for different data storage model configurations (i.e. row or column oriented) and compression techniques can considerably improve the response time of this type of workload for the currently available simulated data of the mission. In addition, we put forward the advantages and shortcomings of the deployment of the framework on a public cloud provider, benchmark against other popular solutions available (that are not always the best for such ad-hoc applications), and describe some user experiences with the framework, which was employed for a number of dedicated astronomical data analysis techniques workshops.

Resilient housing : a new resource-oriented approach

The institutional regimes framework has previously been applied to the institutional conditions that support or hinder the sustainability of housing stocks. This resource-based approach identifies the actors across different sectors that have an interest in housing, how they use housing, the mechanisms affecting their use (public policy, use rights, contracts, etc.) and the effects of their uses on the sustainability of housing within the context of the built environment. The potential of the institutional regimes framework is explored for its suitability to the many considerations of housing resilience. By identifying all the goods and services offered by the resource 'housing stock', researchers and decision-makers could improve the resilience of housing by better accounting for the ecosystem services used by housing, decreasing the vulnerability of housing to disturbances, and maximizing recovery and reorganization following a disturbance. The institutional regimes framework is found to be a promising tool for addressing housing resilience. Further questions are raised for translating this conceptual framework into a practical application underpinned with empirical data.

Description of a robotics-oriented relational positioning methodology

This paper presents a relational positioning methodology for flexibly and intuitively specifying offline programmed robot tasks, as well as for assisting the execution of teleoperated tasks demanding precise movements.In relational positioning, the movements of an object can be restricted totally or partially by specifying its allowed positions in terms of a set of geometric constraints. These allowed positions are found by means of a 3D sequential geometric constraint solver called PMF – Positioning Mobile with respect to Fixed. PMF exploits the fact that in a set of geometric constraints, the rotational component can often be separated from the translational one and solved independently.

Implementing Microsoft .NET Framework for Bitumen-Extranet

Ohjelmistoprojektit pohjautuvat nykyään useasti osittain itsenäisesti suunniteltujen ja; toteutettujen ohjelmakomponenttien yhdistelemiseen. Tällä keinolla voidaan vähentää kehitystyön; viemää aikaa ja kustannuksia, jotta saadaan tuotettua kilpailukykyisempiä ohjelmistoja.; Tässädokumentissa käsitellään komponenttipohjaisen ohjelmistotuotannon näkökulmia ja; Microsoft .NET Framework ympäristöä, joka on kehitysympäristö komponenttipohjaisille; ohjelmistoille. Lisäksi esitellään tapauskohtainen ohjelmistoprojekti extranet-verkon; toteutukseen.

Ohjelmistoyritysten liiketoimintamallit Kaakkois-Suomessa

Tutkielman tavoitteena on tunnistaa Kaakkois-Suomen alueella olevat ohjelmistoyritysten tyypilliset yritysryhmät ja kuvata niiden toimintaa. Tunnistamalla alueen ohjelmistoyrityksille ominaiset piirteet työ antaa myös pohjaa tulevien kehityskohteiden löytämisessä ja kehitystoimenpiteiden kohdistamisessa useisiin samantyyppisiin yrityksiin. Työn taustaksi esitellään ohjelmistoalaa ja ohjelmistoliiketoiminnan malleja, joiden pohjalta muodostetaan viitekehys alueen ohjelmistoyritysten empiiriseen tarkasteluun. Empiriaosuudessa tarkastellaan työn teoreettisessa osiossa esitettyjen liiketoimintamallien toteutumista Kaakkois-Suomessa ja ryhmitellään alueen ohjelmistoyritykset erottelevimpien tekijöiden avulla. Tutkimus on luonteeltaan kvantitatiivinen kokonaistutkimus Kaakkois-Suomen ohjelmistoyrityksistä, ja tutkimusotteeltaan deskriptiivinen eli kuvaileva. Tutkimusaineisto perustui tutkimusryhmän suorittamaan strukturoituun haastatteluun, jossa haastateltiin kaikkiaan 58 ohjelmistoyrityksen vastuuhenkilöitä. Tutkimustulosten perusteella alueelta pystyttiin tunnistamaan neljä toimintatavoiltaan erilaista ohjelmistoliiketoiminnan perustyyppiä: asiakaslähtöiset toimijat (26 toimipaikkaa), räätälöijät (14 toimipaikkaa), integroijat (10 toimipaikkaa) ja tuotteistajat (8 toimipaikkaa). Tulokset osoittavat, että perinteisten ohjelmistoalan liiketoimintamallien kuvaukset ja niistä tehtävät yleistykset antavat hyvän lähtökohdan ohjelmistoyritysten tarkasteluun. Kuitenkin perinteisten ohjelmistoalan liiketoimintamallien antama näkökulma on liian rajoittunut,jos halutaan tarkastella syvällisemmin ohjelmistoyritysten liiketoimintalogiikkaa.

A New Look at Emotional Intelligence: A Dual-Process Framework

In this article, the author provides a framework to guide¦research in emotional intelligence. Studies conducted up¦to the present bear on a conception of emotional intelligence¦as pertaining to the domain of consciousness and¦investigate the construct with a correlational approach.¦As an alternative, the author explores processes underlying¦emotional intelligence, introducing the distinction¦between conscious and automatic processing as a potential¦source of variability in emotionally intelligent¦behavior. Empirical literature is reviewed to support the¦central hypothesis that individual differences in emotional¦intelligence may be best understood by considering¦the way individuals automatically process emotional¦stimuli. Providing directions for research, the author¦encourages the integration of experimental investigation¦of processes underlying emotional intelligence with¦correlational analysis of individual differences and¦fosters the exploration of the automaticity component¦of emotional intelligence.

The role of auditory cortices in the retrieval of single-trial auditory-visual object memories.

Single-trial encounters with multisensory stimuli affect both memory performance and early-latency brain responses to visual stimuli. Whether and how auditory cortices support memory processes based on single-trial multisensory learning is unknown and may differ qualitatively and quantitatively from comparable processes within visual cortices due to purported differences in memory capacities across the senses. We recorded event-related potentials (ERPs) as healthy adults (n = 18) performed a continuous recognition task in the auditory modality, discriminating initial (new) from repeated (old) sounds of environmental objects. Initial presentations were either unisensory or multisensory; the latter entailed synchronous presentation of a semantically congruent or a meaningless image. Repeated presentations were exclusively auditory, thus differing only according to the context in which the sound was initially encountered. Discrimination abilities (indexed by d') were increased for repeated sounds that were initially encountered with a semantically congruent image versus sounds initially encountered with either a meaningless or no image. Analyses of ERPs within an electrical neuroimaging framework revealed that early stages of auditory processing of repeated sounds were affected by prior single-trial multisensory contexts. These effects followed from significantly reduced activity within a distributed network, including the right superior temporal cortex, suggesting an inverse relationship between brain activity and behavioural outcome on this task. The present findings demonstrate how auditory cortices contribute to long-term effects of multisensory experiences on auditory object discrimination. We propose a new framework for the efficacy of multisensory processes to impact both current multisensory stimulus processing and unisensory discrimination abilities later in time.

Project Success through Effective Decisions: Case Studies on Project Goal Setting, Success Evaluation and Managerial Decision Making

The objective of the thesis is to structure and model the factors that contribute to and can be used in evaluating project success. The purpose of this thesis is to enhance the understanding of three research topics. The goal setting process, success evaluation and decision-making process are studied in the context of a project, business unitand its business environment. To achieve the objective three research questionsare posed. These are 1) how to set measurable project goals, 2) how to evaluateproject success and 3) how to affect project success with managerial decisions.The main theoretical contribution comes from deriving a synthesis of these research topics which have mostly been discussed apart from each other in prior research. The research strategy of the study has features from at least the constructive, nomothetical, and decision-oriented research approaches. This strategy guides the theoretical and empirical part of the study. Relevant concepts and a framework are composed on the basis of the prior research contributions within the problem area. A literature review is used to derive constructs of factors withinthe framework. They are related to project goal setting, success evaluation, and decision making. On the basis of this, the case study method is applied to complement the framework. The empirical data includes one product development program, three construction projects, as well as one organization development, hardware/software, and marketing project in their contexts. In two of the case studiesthe analytic hierarchy process is used to formulate a hierarchical model that returns a numerical evaluation of the degree of project success. It has its origin in the solution idea which in turn has its foundation in the notion of projectsuccess. The achieved results are condensed in the form of a process model thatintegrates project goal setting, success evaluation and decision making. The process of project goal setting is analysed as a part of an open system that includes a project, the business unit and its competitive environment. Four main constructs of factors are suggested. First, the project characteristics and requirements are clarified. The second and the third construct comprise the components of client/market segment attractiveness and sources of competitive advantage. Together they determine the competitive position of a business unit. Fourth, the relevant goals and the situation of a business unit are clarified to stress their contribution to the project goals. Empirical evidence is gained on the exploitation of increased knowledge and on the reaction to changes in the business environment during a project to ensure project success. The relevance of a successful project to a company or a business unit tends to increase the higher the reference level of project goals is set. However, normal performance or sometimes performance below this normal level is intentionally accepted. Success measures make project success quantifiable. There are result-oriented, process-oriented and resource-oriented success measures. The study also links result measurements to enablers that portray the key processes. The success measures can be classified into success domains determining the areas on which success is assessed. Empiricalevidence is gained on six success domains: strategy, project implementation, product, stakeholder relationships, learning situation and company functions. However, some project goals, like safety, can be assessed using success measures that belong to two success domains. For example a safety index is used for assessing occupational safety during a project, which is related to project implementation. Product safety requirements, in turn, are connected to the product characteristics and thus to the product-related success domain. Strategic success measures can be used to weave the project phases together. Empirical evidence on their static nature is gained. In order-oriented projects the project phases are oftencontractually divided into different suppliers or contractors. A project from the supplier's perspective can represent only a part of the ¿whole project¿ viewed from the client's perspective. Therefore static success measures are mostly used within the contractually agreed project scope and duration. Proof is also acquired on the dynamic use of operational success measures. They help to focus on the key issues during each project phase. Furthermore, it is shown that the original success domains and success measures, their weights and target values can change dynamically. New success measures can replace the old ones to correspond better with the emphasis of the particular project phase. This adjustment concentrates on the key decision milestones. As a conclusion, the study suggests a combination of static and dynamic success measures. Their linkage to an incentive system can make the project management proactive, enable fast feedback and enhancethe motivation of the personnel. It is argued that the sequence of effective decisions is closely linked to the dynamic control of project success. According to the used definition, effective decisions aim at adequate decision quality and decision implementation. The findings support that project managers construct and use a chain of key decision milestones to evaluate and affect success during aproject. These milestones can be seen as a part of the business processes. Different managers prioritise the key decision milestones to a varying degree. Divergent managerial perspectives, power, responsibilities and involvement during a project offer some explanation for this. Finally, the study introduces the use ofHard Gate and Soft Gate decision milestones. The managers may use the former milestones to provide decision support on result measurements and ad hoc critical conditions. In the latter milestones they may make intermediate success evaluation also on the basis of other types of success measures, like process and resource measures.

On Effort Estimation in Software Projects

Software engineering is criticized as not being engineering or 'well-developed' science at all. Software engineers seem not to know exactly how long their projects will last, what they will cost, and will the software work properly after release. Measurements have to be taken in software projects to improve this situation. It is of limited use to only collect metrics afterwards. The values of the relevant metrics have to be predicted, too. The predictions (i.e. estimates) form the basis for proper project management. One of the most painful problems in software projects is effort estimation. It has a clear and central effect on other project attributes like cost and schedule, and to product attributes like size and quality. Effort estimation can be used for several purposes. In this thesis only the effort estimation in software projects for project management purposes is discussed. There is a short introduction to the measurement issues, and some metrics relevantin estimation context are presented. Effort estimation methods are covered quite broadly. The main new contribution in this thesis is the new estimation model that has been created. It takes use of the basic concepts of Function Point Analysis, but avoids the problems and pitfalls found in the method. It is relativelyeasy to use and learn. Effort estimation accuracy has significantly improved after taking this model into use. A major innovation related to the new estimationmodel is the identified need for hierarchical software size measurement. The author of this thesis has developed a three level solution for the estimation model. All currently used size metrics are static in nature, but this new proposed metric is dynamic. It takes use of the increased understanding of the nature of the work as specification and design work proceeds. It thus 'grows up' along with software projects. The effort estimation model development is not possible without gathering and analyzing history data. However, there are many problems with data in software engineering. A major roadblock is the amount and quality of data available. This thesis shows some useful techniques that have been successful in gathering and analyzing the data needed. An estimation process is needed to ensure that methods are used in a proper way, estimates are stored, reported and analyzed properly, and they are used for project management activities. A higher mechanism called measurement framework is also introduced shortly. The purpose of the framework is to define and maintain a measurement or estimationprocess. Without a proper framework, the estimation capability of an organization declines. It requires effort even to maintain an achieved level of estimationaccuracy. Estimation results in several successive releases are analyzed. It isclearly seen that the new estimation model works and the estimation improvementactions have been successful. The calibration of the hierarchical model is a critical activity. An example is shown to shed more light on the calibration and the model itself. There are also remarks about the sensitivity of the model. Finally, an example of usage is shown.

A logic programming framework for possibilistic argumentation: formalization and logical properties

In the last decade defeasible argumentation frameworks have evolved to become a sound setting to formalize commonsense, qualitative reasoning. The logic programming paradigm has shown to be particularly useful for developing different argument-based frameworks on the basis of different variants of logic programming which incorporate defeasible rules. Most of such frameworks, however, are unable to deal with explicit uncertainty, nor with vague knowledge, as defeasibility is directly encoded in the object language. This paper presents Possibilistic Logic Programming (P-DeLP), a new logic programming language which combines features from argumentation theory and logic programming, incorporating as well the treatment of possibilistic uncertainty. Such features are formalized on the basis of PGL, a possibilistic logic based on G¨odel fuzzy logic. One of the applications of P-DeLP is providing an intelligent agent with non-monotonic, argumentative inference capabilities. In this paper we also provide a better understanding of such capabilities by defining two non-monotonic operators which model the expansion of a given program P by adding new weighed facts associated with argument conclusions and warranted literals, respectively. Different logical properties for the proposed operators are studied