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.

Avian erythrocytes have functional mitochondria, opening novel perspectives for birds as animal models in the study of ageing.

BACKGROUND: In contrast to mammalian erythrocytes, which have lost their nucleus and mitochondria during maturation, the erythrocytes of almost all other vertebrate species are nucleated throughout their lifespan. Little research has been done however to test for the presence and functionality of mitochondria in these cells, especially for birds. Here, we investigated those two points in erythrocytes of one common avian model: the zebra finch (Taeniopygia guttata). RESULTS: Transmission electron microscopy showed the presence of mitochondria in erythrocytes of this small passerine bird, especially after removal of haemoglobin interferences. High-resolution respirometry revealed increased or decreased rates of oxygen consumption by erythrocytes in response to the addition of respiratory chain substrates or inhibitors, respectively. Fluorometric assays confirmed the production of mitochondrial superoxide by avian erythrocytes. Interestingly, measurements of plasmatic oxidative markers indicated lower oxidative stress in blood of the zebra finch compared to a size-matched mammalian model, the mouse. CONCLUSIONS: Altogether, those findings demonstrate that avian erythrocytes possess functional mitochondria in terms of respiratory activities and reactive oxygen species (ROS) production. Interestingly, since blood oxidative stress was lower for our avian model compared to a size-matched mammalian, our results also challenge the idea that mitochondrial ROS production could have been one actor leading to this loss during the course of evolution. Opportunities to assess mitochondrial functioning in avian erythrocytes open new perspectives in the use of birds as models for longitudinal studies of ageing via lifelong blood sampling of the same subjects.

Downregulation of cannabinoid receptor 1 from neuropeptide Y interneurons in the basal ganglia of patients with Huntington's disease and mouse models.

Cannabinoid receptor 1 (CB(1) receptor) controls several neuronal functions, including neurotransmitter release, synaptic plasticity, gene expression and neuronal viability. Downregulation of CB(1) expression in the basal ganglia of patients with Huntington's disease (HD) and animal models represents one of the earliest molecular events induced by mutant huntingtin (mHtt). This early disruption of neuronal CB(1) signaling is thought to contribute to HD symptoms and neurodegeneration. Here we determined whether CB(1) downregulation measured in patients with HD and mouse models was ubiquitous or restricted to specific striatal neuronal subpopulations. Using unbiased semi-quantitative immunohistochemistry, we confirmed previous studies showing that CB(1) expression is downregulated in medium spiny neurons of the indirect pathway, and found that CB(1) is also downregulated in neuropeptide Y (NPY)/neuronal nitric oxide synthase (nNOS)-expressing interneurons while remaining unchanged in parvalbumin- and calretinin-expressing interneurons. CB(1) downregulation in striatal NPY/nNOS-expressing interneurons occurs in R6/2 mice, Hdh(Q150/Q150) mice and the caudate nucleus of patients with HD. In R6/2 mice, CB(1) downregulation in NPY/nNOS-expressing interneurons correlates with diffuse expression of mHtt in the soma. This downregulation also occludes the ability of cannabinoid agonists to activate the pro-survival signaling molecule cAMP response element-binding protein in NPY/nNOS-expressing interneurons. Loss of CB(1) signaling in NPY/nNOS-expressing interneurons could contribute to the impairment of basal ganglia functions linked to HD.

An active contour-based atlas registration model applied to automatic subthalamic nucleus targeting on MRI: method and validation.

This paper presents a new non parametric atlas registration framework, derived from the optical flow model and the active contour theory, applied to automatic subthalamic nucleus (STN) targeting in deep brain stimulation (DBS) surgery. In a previous work, we demonstrated that the STN position can be predicted based on the position of surrounding visible structures, namely the lateral and third ventricles. A STN targeting process can thus be obtained by registering these structures of interest between a brain atlas and the patient image. Here we aim to improve the results of the state of the art targeting methods and at the same time to reduce the computational time. Our simultaneous segmentation and registration model shows mean STN localization errors statistically similar to the most performing registration algorithms tested so far and to the targeting expert's variability. Moreover, the computational time of our registration method is much lower, which is a worthwhile improvement from a clinical point of view.

Transposing phytochrome into the nucleus.

To control many physiological responses, phytochromes directly modulate gene expression. A key regulatory event in this signal transduction pathway is the light-controlled translocation of the photoreceptor from the cytoplasm into the nucleus. Recent publications are beginning to shed light on the molecular mechanisms underlying this central control point. Interestingly, there is a specific mechanism for phytochrome A (phyA) nuclear accumulation. The dedicated phyA nuclear import pathway might be important for the distinct photosensory specificity of this atypical phytochrome. Recent studies in the field also provide a starting point for investigating how the different subcellular pools of phytochrome can control distinct responses to light.

Very high-resolution digital elevation models: are multi-scale-derived variables ecologically relevant?

1. Digital elevation models (DEMs) are often used in landscape ecology to retrieve elevation or first derivative terrain attributes such as slope or aspect in the context of species distribution modelling. However, DEM-derived variables are scale-dependent and, given the increasing availability of very high-resolution (VHR) DEMs, their ecological relevancemust be assessed for different spatial resolutions. 2. In a study area located in the Swiss Western Alps, we computed VHR DEMs-derived variables related to morphometry, hydrology and solar radiation. Based on an original spatial resolution of 0.5 m, we generated DEM-derived variables at 1, 2 and 4 mspatial resolutions, applying a Gaussian Pyramid. Their associations with local climatic factors, measured by sensors (direct and ambient air temperature, air humidity and soil moisture) as well as ecological indicators derived fromspecies composition, were assessed with multivariate generalized linearmodels (GLM) andmixed models (GLMM). 3. Specific VHR DEM-derived variables showed significant associations with climatic factors. In addition to slope, aspect and curvature, the underused wetness and ruggedness indices modelledmeasured ambient humidity and soilmoisture, respectively. Remarkably, spatial resolution of VHR DEM-derived variables had a significant influence on models' strength, with coefficients of determination decreasing with coarser resolutions or showing a local optimumwith a 2 mresolution, depending on the variable considered. 4. These results support the relevance of using multi-scale DEM variables to provide surrogates for important climatic variables such as humidity, moisture and temperature, offering suitable alternatives to direct measurements for evolutionary ecology studies at a local scale.

Species distribution models reveal apparent competitive and facilitative effects of a dominant species on the distribution of tundra plants

Abiotic factors are considered strong drivers of species distribution and assemblages. Yet these spatial patterns are also influenced by biotic interactions. Accounting for competitors or facilitators may improve both the fit and the predictive power of species distribution models (SDMs). We investigated the influence of a dominant species, Empetrum nigrum ssp. hermaphroditum, on the distribution of 34 subordinate species in the tundra of northern Norway. We related SDM parameters of those subordinate species to their functional traits and their co-occurrence patterns with E. hermaphroditum across three spatial scales. By combining both approaches, we sought to understand whether these species may be limited by competitive interactions and/or benefit from habitat conditions created by the dominant species. The model fit and predictive power increased for most species when the frequency of occurrence of E. hermaphroditum was included in the SDMs as a predictor. The largest increase was found for species that 1) co-occur most of the time with E. hermaphroditum, both at large (i.e. 750 m) and small spatial scale (i.e. 2 m) or co-occur with E. hermaphroditum at large scale but not at small scale and 2) have particularly low or high leaf dry matter content (LDMC). Species that do not co-occur with E. hermaphroditum at the smallest scale are generally palatable herbaceous species with low LDMC, thus showing a weak ability to tolerate resource depletion that is directly or indirectly induced by E. hermaphroditum. Species with high LDMC, showing a better aptitude to face resource depletion and grazing, are often found in the proximity of E. hermaphroditum. Our results are consistent with previous findings that both competition and facilitation structure plant distribution and assemblages in the Arctic tundra. The functional and co-occurrence approaches used were complementary and provided a deeper understanding of the observed patterns by refinement of the pool of potential direct and indirect ecological effects of E. hermaphroditum on the distribution of subordinate species. Our correlative study would benefit being complemented by experimental approaches.

Physiotherapy practice in the private sector: organizational characteristics and models.

BACKGROUND: Even if a large proportion of physiotherapists work in the private sector worldwide, very little is known of the organizations within which they practice. Such knowledge is important to help understand contexts of practice and how they influence the quality of services and patient outcomes. The purpose of this study was to: 1) describe characteristics of organizations where physiotherapists practice in the private sector, and 2) explore the existence of a taxonomy of organizational models. METHODS: This was a cross-sectional quantitative survey of 236 randomly-selected physiotherapists. Participants completed a purpose-designed questionnaire online or by telephone, covering organizational vision, resources, structures and practices. Organizational characteristics were analyzed descriptively, while organizational models were identified by multiple correspondence analyses. RESULTS: Most organizations were for-profit (93.2%), located in urban areas (91.5%), and within buildings containing multiple businesses/organizations (76.7%). The majority included multiple providers (89.8%) from diverse professions, mainly physiotherapy assistants (68.7%), massage therapists (67.3%) and osteopaths (50.2%). Four organizational models were identified: 1) solo practice, 2) middle-scale multiprovider, 3) large-scale multiprovider and 4) mixed. CONCLUSIONS: The results of this study provide a detailed description of the organizations where physiotherapists practice, and highlight the importance of human resources in differentiating organizational models. Further research examining the influences of these organizational characteristics and models on outcomes such as physiotherapists' professional practices and patient outcomes are needed.

Modeling sequencing errors by combining Hidden Markov models.

Among the largest resources for biological sequence data is the large amount of expressed sequence tags (ESTs) available in public and proprietary databases. ESTs provide information on transcripts but for technical reasons they often contain sequencing errors. Therefore, when analyzing EST sequences computationally, such errors must be taken into account. Earlier attempts to model error prone coding regions have shown good performance in detecting and predicting these while correcting sequencing errors using codon usage frequencies. In the research presented here, we improve the detection of translation start and stop sites by integrating a more complex mRNA model with codon usage bias based error correction into one hidden Markov model (HMM), thus generalizing this error correction approach to more complex HMMs. We show that our method maintains the performance in detecting coding sequences.

Phenotypic consequences of copy number variation: insights from Smith-Magenis and Potocki-Lupski syndrome mouse models.

A large fraction of genome variation between individuals is comprised of submicroscopic copy number variation of genomic DNA segments. We assessed the relative contribution of structural changes and gene dosage alterations on phenotypic outcomes with mouse models of Smith-Magenis and Potocki-Lupski syndromes. We phenotyped mice with 1n (Deletion/+), 2n (+/+), 3n (Duplication/+), and balanced 2n compound heterozygous (Deletion/Duplication) copies of the same region. Parallel to the observations made in humans, such variation in gene copy number was sufficient to generate phenotypic consequences: in a number of cases diametrically opposing phenotypes were associated with gain versus loss of gene content. Surprisingly, some neurobehavioral traits were not rescued by restoration of the normal gene copy number. Transcriptome profiling showed that a highly significant propensity of transcriptional changes map to the engineered interval in the five assessed tissues. A statistically significant overrepresentation of the genes mapping to the entire length of the engineered chromosome was also found in the top-ranked differentially expressed genes in the mice containing rearranged chromosomes, regardless of the nature of the rearrangement, an observation robust across different cell lineages of the central nervous system. Our data indicate that a structural change at a given position of the human genome may affect not only locus and adjacent gene expression but also "genome regulation." Furthermore, structural change can cause the same perturbation in particular pathways regardless of gene dosage. Thus, the presence of a genomic structural change, as well as gene dosage imbalance, contributes to the ultimate phenotype.

Divergent and convergent margins in the north-western Alps - Confrontation to actualistic models

Divergent and convergent margins actualistic models are reviewed and applied to the history of the western Alps. Tethyan rifting history and geometry are analyzed: the northern European margin is considered as an upper plate whereas the southern Apulian margin is a lower plate; the Breche basin is regarded as the former break-away trough; the internal Brianconnais domain represents the northern rift shoulder whilst the more external domains are regarded as the infill of a complex rim basin locally affected by important extension (Valaisan and Vocontain trough). The Schistes lustres and ophiolites of the Tsate nappe are compared to an accretionary prism: the imbrication of this nappe elements is regarded as a direct consequence of the accretionary phenomena already active in early Cretaceous; the Gets/Simme complex could orginate from a more internal part of the accretionary prism. Some eclogitic basements represent the former Apulian margin substratum (Sesia) others (Mont-Rose) are interpreted as the former edge of the European margin. The history of the closing Tethyan domain is analyzed and the remaining problems concerning the cinematics, the presence/absence of a volcanic arc and the eoalpine metamorphism are discussed.

Automatic Mallampati Classification Using Active Appearance Models

Difficult tracheal intubation assessment is an important research topic in anesthesia as failed intubations are important causes of mortality in anesthetic practice. The modified Mallampati score is widely used, alone or in conjunction with other criteria, to predict the difficulty of intubation. This work presents an automatic method to assess the modified Mallampati score from an image of a patient with the mouth wide open. For this purpose we propose an active appearance models (AAM) based method and use linear support vector machines (SVM) to select a subset of relevant features obtained using the AAM. This feature selection step proves to be essential as it improves drastically the performance of classification, which is obtained using SVM with RBF kernel and majority voting. We test our method on images of 100 patients undergoing elective surgery and achieve 97.9% accuracy in the leave-one-out crossvalidation test and provide a key element to an automatic difficult intubation assessment system.