Advanced Kernel Methods For Remote Sensing Image Classification

Résumé Suite aux recentes avancées technologiques, les archives d'images digitales ont connu une croissance qualitative et quantitative sans précédent. Malgré les énormes possibilités qu'elles offrent, ces avancées posent de nouvelles questions quant au traitement des masses de données saisies. Cette question est à la base de cette Thèse: les problèmes de traitement d'information digitale à très haute résolution spatiale et/ou spectrale y sont considérés en recourant à des approches d'apprentissage statistique, les méthodes à noyau. Cette Thèse étudie des problèmes de classification d'images, c'est à dire de catégorisation de pixels en un nombre réduit de classes refletant les propriétés spectrales et contextuelles des objets qu'elles représentent. L'accent est mis sur l'efficience des algorithmes, ainsi que sur leur simplicité, de manière à augmenter leur potentiel d'implementation pour les utilisateurs. De plus, le défi de cette Thèse est de rester proche des problèmes concrets des utilisateurs d'images satellite sans pour autant perdre de vue l'intéret des méthodes proposées pour le milieu du machine learning dont elles sont issues. En ce sens, ce travail joue la carte de la transdisciplinarité en maintenant un lien fort entre les deux sciences dans tous les développements proposés. Quatre modèles sont proposés: le premier répond au problème de la haute dimensionalité et de la redondance des données par un modèle optimisant les performances en classification en s'adaptant aux particularités de l'image. Ceci est rendu possible par un système de ranking des variables (les bandes) qui est optimisé en même temps que le modèle de base: ce faisant, seules les variables importantes pour résoudre le problème sont utilisées par le classifieur. Le manque d'information étiquétée et l'incertitude quant à sa pertinence pour le problème sont à la source des deux modèles suivants, basés respectivement sur l'apprentissage actif et les méthodes semi-supervisées: le premier permet d'améliorer la qualité d'un ensemble d'entraînement par interaction directe entre l'utilisateur et la machine, alors que le deuxième utilise les pixels non étiquetés pour améliorer la description des données disponibles et la robustesse du modèle. Enfin, le dernier modèle proposé considère la question plus théorique de la structure entre les outputs: l'intègration de cette source d'information, jusqu'à présent jamais considérée en télédétection, ouvre des nouveaux défis de recherche. Advanced kernel methods for remote sensing image classification Devis Tuia Institut de Géomatique et d'Analyse du Risque September 2009 Abstract The technical developments in recent years have brought the quantity and quality of digital information to an unprecedented level, as enormous archives of satellite images are available to the users. However, even if these advances open more and more possibilities in the use of digital imagery, they also rise several problems of storage and treatment. The latter is considered in this Thesis: the processing of very high spatial and spectral resolution images is treated with approaches based on data-driven algorithms relying on kernel methods. In particular, the problem of image classification, i.e. the categorization of the image's pixels into a reduced number of classes reflecting spectral and contextual properties, is studied through the different models presented. The accent is put on algorithmic efficiency and the simplicity of the approaches proposed, to avoid too complex models that would not be used by users. The major challenge of the Thesis is to remain close to concrete remote sensing problems, without losing the methodological interest from the machine learning viewpoint: in this sense, this work aims at building a bridge between the machine learning and remote sensing communities and all the models proposed have been developed keeping in mind the need for such a synergy. Four models are proposed: first, an adaptive model learning the relevant image features has been proposed to solve the problem of high dimensionality and collinearity of the image features. This model provides automatically an accurate classifier and a ranking of the relevance of the single features. The scarcity and unreliability of labeled. information were the common root of the second and third models proposed: when confronted to such problems, the user can either construct the labeled set iteratively by direct interaction with the machine or use the unlabeled data to increase robustness and quality of the description of data. Both solutions have been explored resulting into two methodological contributions, based respectively on active learning and semisupervised learning. Finally, the more theoretical issue of structured outputs has been considered in the last model, which, by integrating outputs similarity into a model, opens new challenges and opportunities for remote sensing image processing.

Delinquents in Iowa Arrest, Services and Sanctions, April 2006

This report was developed to provide summary information to allow practitioners and juvenile justice system officials access to specific sections of Iowa’s Three Year Plan. It includes the “System Flow, “Crime Analysis”, and “Child in Needs of Assistance” sections of Iowa’s 2006 Juvenile Justice and Delinquency Prevention Act formula grant Three-Year Plan. The complete Three Year Plan serves as Iowa’s application for Juvenile Justice and Delinquency Prevention Act formula grant funding. The information included in this report overviews system processing for delinquent youth. It also provides data and analysis from key system decision pointsand services.

Evaluation of Temporary Ramp Metering for Work Zones, 2012

Ramp metering has been successfully implemented in many states to improve traffic operations on freeways. Studies have documented the positive mobility and safety benefits of ramp metering. However, there have been no studies on the use of ramp metering for work zones. This report documents the results from the first deployment of temporary ramp meters in work zones in the United States. Temporary ramp meters were deployed at seven urban short-term work zones in Missouri. Safety measures such as driver compliance, merging behavior, and speed differentials were extracted from video-based field data. Mobility analysis was conducted using a calibrated simulation model and the total delays were obtained for under capacity, at capacity, and over capacity conditions. This evaluation suggests that temporary ramp meters should only be deployed at work zone locations where there is potential for congestion and turned on only during above-capacity conditions. The compliance analysis showed that non-compliance could be a major safety issue in the deployment of temporary ramp meters for under-capacity conditions. The use of a three-section instead of a traditional two-section signal head used for permanent ramp metering produced significantly higher compliance rates. Ramp metering decreased ramp platoons by increasing the percentage of single-vehicle merges to over 70% from under 50%. The accepted-merge-headway results were not statistically significant even though a slight shift towards longer headways was found with the use of ramp meters. Mobility analysis revealed that ramp metering produced delay savings for both mainline and ramp vehicles for work zones operating above capacity. On average a 24% decrease in total delay (mainline plus ramp) at low truck percentage and a 19% decrease in delay at high truck percentage conditions resulted from ramp metering.

Indicadores bibliométricos e econométricos para a avaliação de instituições científicas

O artigo é dividido em duas partes. A primeira descreve a avaliação de uma pequena universidade, baseada em dados cienciométricos tendo por principal objetivo avaliar a pesquisa com visibilidade internacional. A segunda parte mostra como um método econométrico (DEA: data enveloping analysis) pode ser usado para incluir na avaliação o ensino e o levantamento de recursos, entre outros aspectos. As duas abordagens mostram como um corpo teórico combinando bibliometria, cienciometria e econometria pode ser aplicado a problemas concretos.

Integration of Bridge Damage Detection Concepts and Components (3 volumes), TR-636, 2013

This work is divided into three volumes: Volume I: Strain-Based Damage Detection; Volume II: Acceleration-Based Damage Detection; Volume III: Wireless Bridge Monitoring Hardware. Volume I: In this work, a previously-developed structural health monitoring (SHM) system was advanced toward a ready-for-implementation system. Improvements were made with respect to automated data reduction/analysis, data acquisition hardware, sensor types, and communication network architecture. The statistical damage-detection tool, control-chart-based damage-detection methodologies, were further investigated and advanced. For the validation of the damage-detection approaches, strain data were obtained from a sacrificial specimen attached to the previously-utilized US 30 Bridge over the South Skunk River (in Ames, Iowa), which had simulated damage,. To provide for an enhanced ability to detect changes in the behavior of the structural system, various control chart rules were evaluated. False indications and true indications were studied to compare the damage detection ability in regard to each methodology and each control chart rule. An autonomous software program called Bridge Engineering Center Assessment Software (BECAS) was developed to control all aspects of the damage detection processes. BECAS requires no user intervention after initial configuration and training. Volume II: In this work, a previously developed structural health monitoring (SHM) system was advanced toward a ready-for-implementation system. Improvements were made with respect to automated data reduction/analysis, data acquisition hardware, sensor types, and communication network architecture. The objective of this part of the project was to validate/integrate a vibration-based damage-detection algorithm with the strain-based methodology formulated by the Iowa State University Bridge Engineering Center. This report volume (Volume II) presents the use of vibration-based damage-detection approaches as local methods to quantify damage at critical areas in structures. Acceleration data were collected and analyzed to evaluate the relationships between sensors and with changes in environmental conditions. A sacrificial specimen was investigated to verify the damage-detection capabilities and this volume presents a transmissibility concept and damage-detection algorithm that show potential to sense local changes in the dynamic stiffness between points across a joint of a real structure. The validation and integration of the vibration-based and strain-based damage-detection methodologies will add significant value to Iowa’s current and future bridge maintenance, planning, and management Volume III: In this work, a previously developed structural health monitoring (SHM) system was advanced toward a ready-for-implementation system. Improvements were made with respect to automated data reduction/analysis, data acquisition hardware, sensor types, and communication network architecture. This report volume (Volume III) summarizes the energy harvesting techniques and prototype development for a bridge monitoring system that uses wireless sensors. The wireless sensor nodes are used to collect strain measurements at critical locations on a bridge. The bridge monitoring hardware system consists of a base station and multiple self-powered wireless sensor nodes. The base station is responsible for the synchronization of data sampling on all nodes and data aggregation. Each wireless sensor node include a sensing element, a processing and wireless communication module, and an energy harvesting module. The hardware prototype for a wireless bridge monitoring system was developed and tested on the US 30 Bridge over the South Skunk River in Ames, Iowa. The functions and performance of the developed system, including strain data, energy harvesting capacity, and wireless transmission quality, were studied and are covered in this volume.

Contribution of multimorbidity to overall mortality in the population-based CoLaus study

Introduction : Multimorbidity (MM) is currently a major health concern for hospitalized patients but little is known about the relative importance of MM in the general population. Accordingly we assessed whether MM could be a good predictor of overall mortality. Method : Data from the population based CoLaus Study: 3239 participants (1731 women, mean age 50+/-9 years) followed for a median time of 5.4 years (range 0.4 to 8.5 years). MM was defined as presenting >=2 morbidities according to Barnett et al. (27 items, measured data). Survival analysis was conducted using Cox regression. Results : During follow-up, 53 (1.6%) participants died. Participants who died had a higher number of morbidities (2.4 +/- 1.6 vs. 1.9 +/- 1.5, p<0.05) and had a higher prevalence of MM (69.8% vs. 55.9%, p<0.05). On bivariate analysis, presence of MM (defined as a yes/no variable) was significantly related with overall mortality: relative risk (RR) of 1.84, 95% confidence interval [1.02; 3.31], p<0.05 (see figure), but this association became non-significant after adjusting for age, gender and smoking: RR=1.68 [0.93; 3.04], p=0.09. Similar results were obtained when using the number of morbidities: RR for an extra morbidity 1.22 [1.05; 1.44], p<0.02; after adjusting for age, gender and smoking, RR=1.16 [0.99; 1.37], p=0.07. Conclusion : During a short 5 year observation period, measured MM in the general population is associated with overall mortality. This association becomes borderline significant after multivariate adjustment. These observations will have to be confirmed during a longer follow-up period. This increased mortality in MM patients may require developing specific strategies of screening and prevention.

Remorin, a solanaceae protein resident in membrane rafts and plasmodesmata, impairs potato virus X movement.

Remorins (REMs) are proteins of unknown function specific to vascular plants. We have used imaging and biochemical approaches and in situ labeling to demonstrate that REM clusters at plasmodesmata and in approximately 70-nm membrane domains, similar to lipid rafts, in the cytosolic leaflet of the plasma membrane. From a manipulation of REM levels in transgenic tomato (Solanum lycopersicum) plants, we show that Potato virus X (PVX) movement is inversely related to REM accumulation. We show that REM can interact physically with the movement protein TRIPLE GENE BLOCK PROTEIN1 from PVX. Based on the localization of REM and its impact on virus macromolecular trafficking, we discuss the potential for lipid rafts to act as functional components in plasmodesmata and the plasma membrane.

Classification of very high spatial resolution imagery using mathematical morphology and support vector machines

We investigate the relevance of morphological operators for the classification of land use in urban scenes using submetric panchromatic imagery. A support vector machine is used for the classification. Six types of filters have been employed: opening and closing, opening and closing by reconstruction, and opening and closing top hat. The type and scale of the filters are discussed, and a feature selection algorithm called recursive feature elimination is applied to decrease the dimensionality of the input data. The analysis performed on two QuickBird panchromatic images showed that simple opening and closing operators are the most relevant for classification at such a high spatial resolution. Moreover, mixed sets combining simple and reconstruction filters provided the best performance. Tests performed on both images, having areas characterized by different architectural styles, yielded similar results for both feature selection and classification accuracy, suggesting the generalization of the feature sets highlighted.

Statistical significance of quantitative PCR.

BACKGROUND: PCR has the potential to detect and precisely quantify specific DNA sequences, but it is not yet often used as a fully quantitative method. A number of data collection and processing strategies have been described for the implementation of quantitative PCR. However, they can be experimentally cumbersome, their relative performances have not been evaluated systematically, and they often remain poorly validated statistically and/or experimentally. In this study, we evaluated the performance of known methods, and compared them with newly developed data processing strategies in terms of resolution, precision and robustness. RESULTS: Our results indicate that simple methods that do not rely on the estimation of the efficiency of the PCR amplification may provide reproducible and sensitive data, but that they do not quantify DNA with precision. Other evaluated methods based on sigmoidal or exponential curve fitting were generally of both poor resolution and precision. A statistical analysis of the parameters that influence efficiency indicated that it depends mostly on the selected amplicon and to a lesser extent on the particular biological sample analyzed. Thus, we devised various strategies based on individual or averaged efficiency values, which were used to assess the regulated expression of several genes in response to a growth factor. CONCLUSION: Overall, qPCR data analysis methods differ significantly in their performance, and this analysis identifies methods that provide DNA quantification estimates of high precision, robustness and reliability. These methods allow reliable estimations of relative expression ratio of two-fold or higher, and our analysis provides an estimation of the number of biological samples that have to be analyzed to achieve a given precision.

Dental service utilization by Europeans aged 50 plus.

OBJECTIVES: To describe variations in the utilization of dental services by persons aged 50+ from 14 European countries and to identify the extent to which such variations are attributable to differences in oral health need and in accessibility of dental care. METHODS: We use data from the Survey of Health, Ageing, and Retirement in Europe (SHARE Waves 2 and 3) and estimate a series of multivariate logistic regression models to analyze variations in dental service utilization (overall dental attendance, preventive treatment and/or operative treatment, dental attendance in early life years) RESULTS: Overall dental attendance and incidence of solely preventive treatment are comparatively high in the Netherlands, Sweden, Denmark, Germany, and Switzerland. In contrast, overall dental attendance is relatively low in Spain, Italy, France, Greece, Poland, and Ireland. Moreover, a high incidence of solely operative treatment is observed in Austria, Italy, and France, whereas in the Netherlands, Sweden, Denmark, Switzerland, and Ireland, the incidence of solely operative treatment is comparably low. By and large, these variations persist even when controlling for cross-country differences in oral health need and in accessibility of dental care. CONCLUSIONS: In comparison with other European regions, there is a tendency toward more frequent and preventive dental treatment of the elderly populations residing in Scandinavia and Western Europe. Such utilization patterns appear only partially attributable to differences in need for and accessibility of dental care.

Relative Technical Efficiency of European Transportation Systems

Talouden kasvaessa myös tavarankuljetusmäärät kasvavat. Kuljetusjärjestelmät ja niiden sujuva toiminta on erittäin tärkeää taloudellisen kasvun kannalta tällä hetkellä, ja se tulee olemaan yhä tärkeämpää tulevaisuudessa. Tulevaisuudessa tarvitaan kokonaisvaltainen ja selkeästi tehokkaampi kuljetusjärjestelmä, mikäli tulevaisuuden kuljetusvirrat halutaan hoitaa kestävästi. Tässä opinnäytetyössäni tutkin kolmen eurooppalaisen kuljetusjärjestelmän (rautatiet, lentoliikenne ja konttiliikenne meritse) suhteellista teknistä tehokkuutta ja menetelmänä on data envelopment analysis (DEA). Vertailtaessa kuljetusjärjestelmiä löytyi suuria eroja kuljetusmuotojen välille. lentoyhtiöt suoriutuivat huomattavan tasaisesti eli tehokkaiden ja ei-tehokkaiden toimijoiden välillä ei ollut suuria eroja. Rautatiepuolella erot venyivät huomattavan suuriksi niin eri yritysten välillä kuin jopa saman yrityksen sisällä eri vuosina. Pikaisemmassa laivayhtiöiden tarkastelussa erot niiden välillä olivat lähes yhtä pieniä kuin lentoyhtiöiden välillä. Tarkasteltaessa omistajuuden vaikutusta lentoyhtiöiden toiminnassa huomattiin, että yksityisessä omistuksessa olevat yritykset olivat huomattavasti tehokkaampia matkustajien kuljettamisessa. Rahtipuolella merkittäviä eroja ei havaittu. Merkittävät korrelaatiot eri mallien välillä antoivat joitain viitteitä myös kuljetuspoliittiseen päätöksentekoon; investoinnit matkustajienkuljetuksiin raiteilla parantaisivat koko rautatiepuolen teokkuutta, mutta myös samalla lentopuolen matkustajakuljetuksen tehokkuutta.

Ranking RegionalInnovation Policies: DEA-based Benchmarking in an European Setting

Seudullinen innovaatio on monimutkainen ilmiö, joka usein sijaitsee paikallisten toimijoiden keskinäisen vuorovaikutuksen kentässä. Täten sitä on perinteisesti pidetty vaikeasti mitattavana ilmiönä. Työssä sovellettiin Data Envelopment Analysis menetelmää, joka on osoittautunut aiemmin menestyksekkääksi tapauksissa, joissa mitattavien syötteiden ja tuotteiden väliset suhteet eivät ole olleet ilmeisiä. Työssä luotiin konseptuaalinen malli seudullisen innovaation syötteistä ja tuotteista, jonka perusteella valittiin 12 tilastollisen muuttujan mittaristo. Käyttäen Eurostat:ia datalähteenä, lähdedata kahdeksaan muuttujsta saatiin seudullisella tasolla, sekä mittaristoa täydennettiin yhdellä kansallisella muuttujalla. Arviointi suoritettiin lopulta 45 eurooppalaiselle seudulle. Tutkimuksen painopiste oli arvioida DEA-menetelmän soveltuvuutta innovaatio-järjestelmän mittaamiseen, sillä menetelmää ei ole aiemmin sovellettu vastaavassa tapauksessa. Ensimmäiset tulokset osoittivat ylipäätään liiallisen korkeita tehok-kuuslukuja. Korjaustoimenpiteitä erottelutarkkuuden parantamiseksi esiteltiin ja sovellettiin, jonka jälkeen saatiin realistisempia tuloksia ja ranking-lista arvioitavista seuduista. DEA-menetelmän todettiin olevan tehokas ja kiinnostava työkalu arviointikäytäntöjen ja innovaatiopolitiikan kehittämiseen, sikäli kun datan saatavuusongelmat saadaan ratkaistua sekä itse mallia tarkennettua.

A Monte Carlo approach to the inverse problem of diffuse pollution risk in agricultural catchments

The hydrological and biogeochemical processes that operate in catchments influence the ecological quality of freshwater systems through delivery of fine sediment, nutrients and organic matter. Most models that seek to characterise the delivery of diffuse pollutants from land to water are reductionist. The multitude of processes that are parameterised in such models to ensure generic applicability make them complex and difficult to test on available data. Here, we outline an alternative - data-driven - inverse approach. We apply SCIMAP, a parsimonious risk based model that has an explicit treatment of hydrological connectivity. we take a Bayesian approach to the inverse problem of determining the risk that must be assigned to different land uses in a catchment in order to explain the spatial patterns of measured in-stream nutrient concentrations. We apply the model to identify the key sources of nitrogen (N) and phosphorus (P) diffuse pollution risk in eleven UK catchments covering a range of landscapes. The model results show that: 1) some land use generates a consistently high or low risk of diffuse nutrient pollution; but 2) the risks associated with different land uses vary both between catchments and between nutrients; and 3) that the dominant sources of P and N risk in the catchment are often a function of the spatial configuration of land uses. Taken on a case-by-case basis, this type of inverse approach may be used to help prioritise the focus of interventions to reduce diffuse pollution risk for freshwater ecosystems. (C) 2012 Elsevier B.V. All rights reserved.

Eficiencia técnica y convergencia en la industria manufacturera de la Unión Europea

El objetivo de este trabajo es estudiar la evolución de los niveles de eficiencia técnica de los principales sectores de la industria manufacturera europea durante el periodo 1987-1996. Para ello se ha aplicado un análisis envolvente de datos (DEA) con la información obtenida de la base de datos BACH de 1996. Los resultados muestran que la eficiencia media de la industria disminuye en este período. Además, no encontra-mos una evidencia clara de que haya habido convergencia en la eficiencia entre las empresas europeas. No obstante, nuestro análisis revela una relación estrecha del ciclo económico con los niveles de eficiencia y con su dispersión.

Ranking Regional Innovation Policies: DEA-based Benchmarking in an European Setting

Seudullinen innovaatio on monimutkainen ilmiö, joka usein sijaitsee paikallisten toimijoiden keskinäisen vuorovaikutuksen kentässä. Täten sitä on perinteisesti pidetty vaikeasti mitattavana ilmiönä. Työssä sovellettiin Data Envelopment Analysis menetelmää, joka on osoittautunut aiemmin menestyksekkääksi tapauksissa, joissa mitattavien syötteiden ja tuotteiden väliset suhteet eivät ole olleet ilmeisiä. Työssä luotiin konseptuaalinen malli seudullisen innovaation syötteistä ja tuotteista, jonka perusteella valittiin 12 tilastollisen muuttujan mittaristo. Käyttäen Eurostat:ia datalähteenä, lähdedata kahdeksaan muuttujsta saatiin seudullisella tasolla, sekä mittaristoa täydennettiin yhdellä kansallisella muuttujalla. Arviointi suoritettiin lopulta 45 eurooppalaiselle seudulle. Tutkimuksen painopiste oli arvioida DEA-menetelmän soveltuvuutta innovaatiojärjestelmän mittaamiseen, sillä menetelmää ei ole aiemmin sovellettu vastaavassa tapauksessa. Ensimmäiset tulokset osoittivat ylipäätään liiallisen korkeita tehokkuuslukuja. Korjaustoimenpiteitä erottelutarkkuuden parantamiseksi esiteltiin ja sovellettiin, jonka jälkeen saatiin realistisempia tuloksia ja ranking-lista arvioitavista seuduista. DEA-menetelmän todettiin olevan tehokas ja kiinnostava työkalu arviointikäytäntöjen ja innovaatiopolitiikan kehittämiseen, sikäli kun datan saatavuusongelmat saadaan ratkaistua sekä itse mallia tarkennettua.