Geographic assignment of seabirds to their origin: combining morphology, genetics and biogeochemical analyses

Longline fisheries, oil spills, and offshore wind farms are some of the major threats increasing seabird mortality at sea, but the impact of these threats on specific populations has been difficult to determine so far. We tested the use of molecular markers, morphometric measures, and stable isotope (δ15N and δ13C) and trace element concentrations in the first primary feather (grown at the end of the breeding period) to assign the geographic origin of Calonectris shearwaters. Overall, we sampled birds from three taxa: 13 Mediterranean Cory's Shearwater (Calonectris diomedea diomedea) breeding sites, 10 Atlantic Cory's Shearwater (Calonectris diomedea borealis) breeding sites, and one Cape Verde Shearwater (C. edwardsii) breeding site. Assignment rates were investigated at three spatial scales: breeding colony, breeding archipelago, and taxa levels. Genetic analyses based on the mitochondrial control region (198 birds from 21 breeding colonies) correctly assigned 100% of birds to the three main taxa but failed in detecting geographic structuring at lower scales. Discriminant analyses based on trace elements composition achieved the best rate of correct assignment to colony (77.5%). Body measurements or stable isotopes mainly succeeded in assigning individuals among taxa (87.9% and 89.9%, respectively) but failed at the colony level (27.1% and 38.0%, respectively). Combining all three approaches (morphometrics, isotopes, and trace elements on 186 birds from 15 breeding colonies) substantially improved correct classifications (86.0%, 90.7%, and 100% among colonies, archipelagos, and taxa, respectively). Validations using two independent data sets and jackknife cross-validation confirmed the robustness of the combined approach in the colony assignment (62.5%, 58.8%, and 69.8% for each validation test, respectively). A preliminary application of the discriminant model based on stable isotope δ15N and δ13C values and trace elements (219 birds from 17 breeding sites) showed that 41 Cory's Shearwaters caught by western Mediterranean long-liners came mainly from breeding colonies in Menorca (48.8%), Ibiza (14.6%), and Crete (31.7%). Our findings show that combining analyses of trace elements and stable isotopes on feathers can achieve high rates of correct geographic assignment of birds in the marine environment, opening new prospects for the study of seabird mortality at sea.

Color ad spectral image assessment using novel quality and fidelity techniques

The ongoing development of the digital media has brought a new set of challenges with it. As images containing more than three wavelength bands, often called spectral images, are becoming a more integral part of everyday life, problems in the quality of the RGB reproduction from the spectral images have turned into an important area of research. The notion of image quality is often thought to comprise two distinctive areas – image quality itself and image fidelity, both dealing with similar questions, image quality being the degree of excellence of the image, and image fidelity the measure of the match of the image under study to the original. In this thesis, both image fidelity and image quality are considered, with an emphasis on the influence of color and spectral image features on both. There are very few works dedicated to the quality and fidelity of spectral images. Several novel image fidelity measures were developed in this study, which include kernel similarity measures and 3D-SSIM (structural similarity index). The kernel measures incorporate the polynomial, Gaussian radial basis function (RBF) and sigmoid kernels. The 3D-SSIM is an extension of a traditional gray-scale SSIM measure developed to incorporate spectral data. The novel image quality model presented in this study is based on the assumption that the statistical parameters of the spectra of an image influence the overall appearance. The spectral image quality model comprises three parameters of quality: colorfulness, vividness and naturalness. The quality prediction is done by modeling the preference function expressed in JNDs (just noticeable difference). Both image fidelity measures and the image quality model have proven to be effective in the respective experiments.

Comparison Structure Analysis

This study presents an automatic, computer-aided analytical method called Comparison Structure Analysis (CSA), which can be applied to different dimensions of music. The aim of CSA is first and foremost practical: to produce dynamic and understandable representations of musical properties by evaluating the prevalence of a chosen musical data structure through a musical piece. Such a comparison structure may refer to a mathematical vector, a set, a matrix or another type of data structure and even a combination of data structures. CSA depends on an abstract systematic segmentation that allows for a statistical or mathematical survey of the data. To choose a comparison structure is to tune the apparatus to be sensitive to an exclusive set of musical properties. CSA settles somewhere between traditional music analysis and computer aided music information retrieval (MIR). Theoretically defined musical entities, such as pitch-class sets, set-classes and particular rhythm patterns are detected in compositions using pattern extraction and pattern comparison algorithms that are typical within the field of MIR. In principle, the idea of comparison structure analysis can be applied to any time-series type data and, in the music analytical context, to polyphonic as well as homophonic music. Tonal trends, set-class similarities, invertible counterpoints, voice-leading similarities, short-term modulations, rhythmic similarities and multiparametric changes in musical texture were studied. Since CSA allows for a highly accurate classification of compositions, its methods may be applicable to symbolic music information retrieval as well. The strength of CSA relies especially on the possibility to make comparisons between the observations concerning different musical parameters and to combine it with statistical and perhaps other music analytical methods. The results of CSA are dependent on the competence of the similarity measure. New similarity measures for tonal stability, rhythmic and set-class similarity measurements were proposed. The most advanced results were attained by employing the automated function generation – comparable with the so-called genetic programming – to search for an optimal model for set-class similarity measurements. However, the results of CSA seem to agree strongly, independent of the type of similarity function employed in the analysis.

Étude de dispositifs électroniques moléculaires à l'aide de modèles simples

Cette thèse en électronique moléculaire porte essentiellement sur le développement d’une méthode pour le calcul de la transmission de dispositifs électroniques moléculaires (DEMs), c’est-à-dire des molécules branchées à des contacts qui forment un dispositif électronique de taille moléculaire. D’une part, la méthode développée vise à apporter un point de vue différent de celui provenant des méthodes déjà existantes pour ce type de calculs. D’autre part, elle permet d’intégrer de manière rigoureuse des outils théoriques déjà développés dans le but d’augmenter la qualité des calculs. Les exemples simples présentés dans ce travail permettent de mettre en lumière certains phénomènes, tel que l’interférence destructive dans les dispositifs électroniques moléculaires. Les chapitres proviennent d’articles publiés dans la littérature. Au chapitre 2, nous étudions à l’aide d’un modèle fini avec la méthode de la théorie de la fonctionnelle de la densité de Kohn-Sham un point quantique moléculaire. De plus, nous calculons la conductance du point quantique moléculaire avec une implémentation de la formule de Landauer. Nous trouvons que la structure électronique et la conductance moléculaire dépendent fortement de la fonctionnelle d’échange et de corrélation employée. Au chapitre 3, nous discutons de l’effet de l’ajout d’une chaîne ramifiée à des molécules conductrices sur la probabilité de transmission de dispositifs électroniques moléculaires. Nous trouvons que des interférences destructives apparaissent aux valeurs propres de l’énergie des chaînes ramifiées isolées, si ces valeurs ne correspondent pas à des états localisés éloignés du conducteur moléculaire. Au chapitre 4, nous montrons que les dispositifs électroniques moléculaires contenant une molécule aromatique présentent généralement des courants circulaires qui sont associés aux phénomènes d’interférence destructive dans ces systèmes. Au chapitre 5, nous employons l’approche « source-sink potential » (SSP) pour étudier la transmission de dispositifs électroniques moléculaires. Au lieu de considérer les potentiels de sources et de drains exactement, nous utilisons la théorie des perturbations pour trouver une expression de la probabilité de transmission, T(E) = 1 − |r(E)|2, où r(E) est le coefficient de réflexion qui dépend de l’énergie. Cette expression dépend des propriétés de la molécule isolée, en effet nous montrons que c’est la densité orbitalaire sur les atomes de la molécule qui sont connectés aux contacts qui détermine principalement la transmission du dispositif à une énergie de l’électron incident donnée. Au chapitre 6, nous présentons une extension de l’approche SSP à un canal pour des dispositifs électroniques moléculaires à plusieurs canaux. La méthode à multiples canaux proposée repose sur une description des canaux propres des états conducteurs du dispositif électronique moléculaire (DEM) qui sont obtenus par un algorithme auto-cohérent. Finalement, nous utilisons le modèle développé afin d’étudier la transmission du 1-phényl-1,3-butadiène branché à deux rangées d’atomes couplées agissant comme contacts à gauche et à la droite.

Estudi de mètodes de classificació borrosa i la seva aplicació a l'agrupació de zones geogràfiques en base a diverses característiques incertes

Aquesta memòria està estructurada en sis capítols amb l'objectiu final de fonamentar i desenvolupar les eines matemàtiques necessàries per a la classificació de conjunts de subconjunts borrosos. El nucli teòric del treball el formen els capítols 3, 4 i 5; els dos primers són dos capítols de caire més general, i l'últim és una aplicació dels anteriors a la classificació dels països de la Unió Europea en funció de determinades característiques borroses. En el capítol 1 s'analitzen les diferents connectives borroses posant una especial atenció en aquells aspectes que en altres capítols tindran una aplicació específica. És per aquest motiu que s'estudien les ordenacions de famílies de t-normes, donada la seva importància en la transitivitat de les relacions borroses. La verificació del principi del terç exclòs és necessària per assegurar que un conjunt significatiu de mesures borroses generalitzades, introduïdes en el capítol 3, siguin reflexives. Estudiem per a quines t-normes es verifica aquesta propietat i introduïm un nou conjunt de t-normes que verifiquen aquest principi. En el capítol 2 es fa un recorregut general per les relacions borroses centrant-nos en l'estudi de la clausura transitiva per a qualsevol t-norma, el càlcul de la qual és en molts casos fonamental per portar a terme el procés de classificació. Al final del capítol s'exposa un procediment pràctic per al càlcul d'una relació borrosa amb l'ajuda d'experts i de sèries estadístiques. El capítol 3 és un monogràfic sobre mesures borroses. El primer objectiu és relacionar les mesures (o distàncies) usualment utilitzades en les aplicacions borroses amb les mesures conjuntistes crisp. Es tracta d'un enfocament diferent del tradicional enfocament geomètric. El principal resultat és la introducció d'una família parametritzada de mesures que verifiquen unes propietats de caràcter conjuntista prou satisfactòries. L'estudi de la verificació del principi del terç exclòs té aquí la seva aplicació sobre la reflexivitat d'aquestes mesures, que són estudiades amb una certa profunditat en alguns casos particulars. El capítol 4 és, d'entrada, un repàs dels principals resultats i mètodes borrosos per a la classificació dels elements d'un mateix conjunt de subconjunts borrosos. És aquí on s'apliquen els resultats sobre les ordenacions de les famílies de t-normes i t-conormes estudiades en el capítol 1. S'introdueix un nou mètode de clusterització, canviant la matriu de la relació borrosa cada vegada que s'obté un nou clúster. Aquest mètode permet homogeneïtzar la metodologia del càlcul de la relació borrosa amb el mètode de clusterització. El capítol 5 tracta sobre l'agrupació d'objectes de diferent naturalesa; és a dir, subconjunts borrosos que pertanyen a diferents conjunts. Aquesta teoria ja ha estat desenvolupada en el cas binari; aquí, el que es presenta és la seva generalització al cas n-ari. Més endavant s'estudien certs aspectes de les projeccions de la relació sobre un cert espai i el recíproc, l'estudi de cilindres de relacions predeterminades. Una aplicació sobre l'agrupació de les comarques gironines en funció de certes variables borroses es presenta al final del capítol. L'últim capítol és eminentment pràctic, ja que s'aplica allò estudiat principalment en els capítols 3 i 4 a la classificació dels països de la Unió Europea en funció de determinades característiques borroses. Per tal de fer previsions per a anys venidors s'han utilitzat sèries temporals i xarxes neuronals. S'han emprat diverses mesures i mètodes de clusterització per tal de poder comparar els diversos dendogrames que resulten del procés de clusterització. Finalment, als annexos es poden consultar les sèries estadístiques utilitzades, la seva extrapolació, els càlculs per a la construcció de les matrius de les relacions borroses, les matrius de mesura i les seves clausures.

Distributional semantics and machine learning for statistical machine translation

[EU]Lan honetan semantika distribuzionalaren eta ikasketa automatikoaren erabilera aztertzen dugu itzulpen automatiko estatistikoa hobetzeko. Bide horretan, erregresio logistikoan oinarritutako ikasketa automatikoko eredu bat proposatzen dugu hitz-segiden itzulpen- probabilitatea modu dinamikoan modelatzeko. Proposatutako eredua itzulpen automatiko estatistikoko ohiko itzulpen-probabilitateen orokortze bat dela frogatzen dugu, eta testuinguruko nahiz semantika distribuzionaleko informazioa barneratzeko baliatu ezaugarri lexiko, hitz-cluster eta hitzen errepresentazio bektorialen bidez. Horretaz gain, semantika distribuzionaleko ezagutza itzulpen automatiko estatistikoan txertatzeko beste hurbilpen bat lantzen dugu: hitzen errepresentazio bektorial elebidunak erabiltzea hitz-segiden itzulpenen antzekotasuna modelatzeko. Gure esperimentuek proposatutako ereduen baliagarritasuna erakusten dute, emaitza itxaropentsuak eskuratuz oinarrizko sistema sendo baten gainean. Era berean, gure lanak ekarpen garrantzitsuak egiten ditu errepresentazio bektorialen mapaketa elebidunei eta hitzen errepresentazio bektorialetan oinarritutako hitz-segiden antzekotasun neurriei dagokienean, itzulpen automatikoaz haratago balio propio bat dutenak semantika distribuzionalaren arloan.

CID: an efficient complexity-invariant distance for time series

The ubiquity of time series data across almost all human endeavors has produced a great interest in time series data mining in the last decade. While dozens of classification algorithms have been applied to time series, recent empirical evidence strongly suggests that simple nearest neighbor classification is exceptionally difficult to beat. The choice of distance measure used by the nearest neighbor algorithm is important, and depends on the invariances required by the domain. For example, motion capture data typically requires invariance to warping, and cardiology data requires invariance to the baseline (the mean value). Similarly, recent work suggests that for time series clustering, the choice of clustering algorithm is much less important than the choice of distance measure used.In this work we make a somewhat surprising claim. There is an invariance that the community seems to have missed, complexity invariance. Intuitively, the problem is that in many domains the different classes may have different complexities, and pairs of complex objects, even those which subjectively may seem very similar to the human eye, tend to be further apart under current distance measures than pairs of simple objects. This fact introduces errors in nearest neighbor classification, where some complex objects may be incorrectly assigned to a simpler class. Similarly, for clustering this effect can introduce errors by “suggesting” to the clustering algorithm that subjectively similar, but complex objects belong in a sparser and larger diameter cluster than is truly warranted.We introduce the first complexity-invariant distance measure for time series, and show that it generally produces significant improvements in classification and clustering accuracy. We further show that this improvement does not compromise efficiency, since we can lower bound the measure and use a modification of triangular inequality, thus making use of most existing indexing and data mining algorithms. We evaluate our ideas with the largest and most comprehensive set of time series mining experiments ever attempted in a single work, and show that complexity-invariant distance measures can produce improvements in classification and clustering in the vast majority of cases.

Atomic-scale characterization of diamond surfaces and fullerene self-assembly

In this thesis, elemental research towards the implantation of a diamond-based molecular quantum computer is presented. The approach followed requires linear alignment of endohedral fullerenes on the diamond C(100) surface in the vicinity of subsurface NV-centers. From this, four fundamental experimental challenges arise: 1) The well-controlled deposition of endohedral fullerenes on a diamond surface. 2) The creation of NV-centers in diamond close to the surface. 3) Preparation and characterization of atomically-flat diamondsurfaces. 4) Assembly of linear chains of endohedral fullerenes. First steps to overcome all these challenges were taken in the framework of this thesis. Therefore, a so-called “pulse injection” technique was implemented and tested in a UHV chamber that was custom-designed for this and further tasks. Pulse injection in principle allows for the deposition of molecules from solution onto a substrate and can therefore be used to deposit molecular species that are not stable to sublimation under UHV conditions, such as the endohedral fullerenes needed for a quantum register. Regarding the targeted creation of NV-centers, FIB experiments were carried out in cooperation with the group of Prof. Schmidt-Kaler (AG Quantum, Physics Department, Johannes Gutenberg-Universität Mainz). As an entry into this challenging task, argon cations were implanted into (111) surface-oriented CaF2 crystals. The resulting implantation spots on the surface were imaged and characterized using AFM. In this context, general relations between the impact of the ions on the surface and their valency or kinetic energy, respectively, could be established. The main part of this thesis, however, is constituted by NCAFM studies on both, bare and hydrogen-terminated diamond C(100) surfaces. In cooperation with the group of Prof. Dujardin (Molecular Nanoscience Group, ISMO, Université de Paris XI), clean and atomically-flat diamond surfaces were prepared by exposure of the substrate to a microwave hydrogen plasma. Subsequently, both surface modifications were imaged in high resolution with NC-AFM. In the process, both hydrogen atoms in the unit cell of the hydrogenated surface were resolved individually, which was not achieved in previous STM studies of this surface. The NC-AFM images also reveal, for the first time, atomic-resolution contrast on the clean, insulating diamond surface and provide real-space experimental evidence for a (2×1) surface reconstruction. With regard to the quantum computing concept, high-resolution NC-AFM imaging was also used to study the adsorption and self-assembly potential of two different kinds of fullerenes (C60 and C60F48) on aforementioned diamond surfaces. In case of the hydrogenated surface, particular attention was paid to the influence of charge transfer doping on the fullerene-substrate interaction and the morphology emerging from self-assembly. Finally, self-assembled C60 islands on the hydrogen-terminated diamond surface were subject to active manipulation by an NC-AFM tip. Two different kinds of tip-induced island growth modes have been induced and were presented. In conclusion, the results obtained provide fundamental informations mandatory for the realization of a molecular quantum computer. In the process it was shown that NC-AFM is, under proper circumstances, a very capable tool for imaging diamond surfaces with highest resolution, surpassing even what has been achieved with STM up to now. Particular attention was paid to the influence of transfer doping on the morphology of fullerenes on the hydrogenated diamond surface, revealing new possibilities for tailoring the self-assembly of molecules that have a high electron affinity.

Application and assessment of a robust elastic motion correction algorithm to dynamic MRI

The purpose of this study was to assess the performance of a new motion correction algorithm. Twenty-five dynamic MR mammography (MRM) data sets and 25 contrast-enhanced three-dimensional peripheral MR angiographic (MRA) data sets which were affected by patient motion of varying severeness were selected retrospectively from routine examinations. Anonymized data were registered by a new experimental elastic motion correction algorithm. The algorithm works by computing a similarity measure for the two volumes that takes into account expected signal changes due to the presence of a contrast agent while penalizing other signal changes caused by patient motion. A conjugate gradient method is used to find the best possible set of motion parameters that maximizes the similarity measures across the entire volume. Images before and after correction were visually evaluated and scored by experienced radiologists with respect to reduction of motion, improvement of image quality, disappearance of existing lesions or creation of artifactual lesions. It was found that the correction improves image quality (76% for MRM and 96% for MRA) and diagnosability (60% for MRM and 96% for MRA).

Development of a physically-based method for delineation of hydrologically homogeneous regions and flood quantile estimation in ungauged basins via the index flood method

Regional flood frequency techniques are commonly used to estimate flood quantiles when flood data is unavailable or the record length at an individual gauging station is insufficient for reliable analyses. These methods compensate for limited or unavailable data by pooling data from nearby gauged sites. This requires the delineation of hydrologically homogeneous regions in which the flood regime is sufficiently similar to allow the spatial transfer of information. It is generally accepted that hydrologic similarity results from similar physiographic characteristics, and thus these characteristics can be used to delineate regions and classify ungauged sites. However, as currently practiced, the delineation is highly subjective and dependent on the similarity measures and classification techniques employed. A standardized procedure for delineation of hydrologically homogeneous regions is presented herein. Key aspects are a new statistical metric to identify physically discordant sites, and the identification of an appropriate set of physically based measures of extreme hydrological similarity. A combination of multivariate statistical techniques applied to multiple flood statistics and basin characteristics for gauging stations in the Southeastern U.S. revealed that basin slope, elevation, and soil drainage largely determine the extreme hydrological behavior of a watershed. Use of these characteristics as similarity measures in the standardized approach for region delineation yields regions which are more homogeneous and more efficient for quantile estimation at ungauged sites than those delineated using alternative physically-based procedures typically employed in practice. The proposed methods and key physical characteristics are also shown to be efficient for region delineation and quantile development in alternative areas composed of watersheds with statistically different physical composition. In addition, the use of aggregated values of key watershed characteristics was found to be sufficient for the regionalization of flood data; the added time and computational effort required to derive spatially distributed watershed variables does not increase the accuracy of quantile estimators for ungauged sites. This dissertation also presents a methodology by which flood quantile estimates in Haiti can be derived using relationships developed for data rich regions of the U.S. As currently practiced, regional flood frequency techniques can only be applied within the predefined area used for model development. However, results presented herein demonstrate that the regional flood distribution can successfully be extrapolated to areas of similar physical composition located beyond the extent of that used for model development provided differences in precipitation are accounted for and the site in question can be appropriately classified within a delineated region.

MQN-Mapplet: Visualization of Chemical Space with Interactive Maps of DrugBank, ChEMBL, PubChem, GDB-11, and GDB-13

The MQN-mapplet is a Java application giving access to the structure of small molecules in large databases via color-coded maps of their chemical space. These maps are projections from a 42-dimensional property space defined by 42 integer value descriptors called molecular quantum numbers (MQN), which count different categories of atoms, bonds, polar groups, and topological features and categorize molecules by size, rigidity, and polarity. Despite its simplicity, MQN-space is relevant to biological activities. The MQN-mapplet allows localization of any molecule on the color-coded images, visualization of the molecules, and identification of analogs as neighbors on the MQN-map or in the original 42-dimensional MQN-space. No query molecule is necessary to start the exploration, which may be particularly attractive for nonchemists. To our knowledge, this type of interactive exploration tool is unprecedented for very large databases such as PubChem and GDB-13 (almost one billion molecules). The application is freely available for download at www.gdb.unibe.ch.

Generalized fuzzy clustering for segmentation of multi-spectral magnetic resonance images.

An integrated approach for multi-spectral segmentation of MR images is presented. This method is based on the fuzzy c-means (FCM) and includes bias field correction and contextual constraints over spatial intensity distribution and accounts for the non-spherical cluster's shape in the feature space. The bias field is modeled as a linear combination of smooth polynomial basis functions for fast computation in the clustering iterations. Regularization terms for the neighborhood continuity of intensity are added into the FCM cost functions. To reduce the computational complexity, the contextual regularizations are separated from the clustering iterations. Since the feature space is not isotropic, distance measure adopted in Gustafson-Kessel (G-K) algorithm is used instead of the Euclidean distance, to account for the non-spherical shape of the clusters in the feature space. These algorithms are quantitatively evaluated on MR brain images using the similarity measures.

Volume and shape in feature space on adaptive FCM in MRI segmentation.

Intensity non-uniformity (bias field) correction, contextual constraints over spatial intensity distribution and non-spherical cluster's shape in the feature space are incorporated into the fuzzy c-means (FCM) for segmentation of three-dimensional multi-spectral MR images. The bias field is modeled by a linear combination of smooth polynomial basis functions for fast computation in the clustering iterations. Regularization terms for the neighborhood continuity of either intensity or membership are added into the FCM cost functions. Since the feature space is not isotropic, distance measures, other than the Euclidean distance, are used to account for the shape and volumetric effects of clusters in the feature space. The performance of segmentation is improved by combining the adaptive FCM scheme with the criteria used in Gustafson-Kessel (G-K) and Gath-Geva (G-G) algorithms through the inclusion of the cluster scatter measure. The performance of this integrated approach is quantitatively evaluated on normal MR brain images using the similarity measures. The improvement in the quality of segmentation obtained with our method is also demonstrated by comparing our results with those produced by FSL (FMRIB Software Library), a software package that is commonly used for tissue classification.

The Chemical Space of Flavours

The flavour of foods is determined by the interaction of taste molecules with receptors in the mouth, and fragrances or aroma with receptors in the upper part of the nose. Here, we discuss the properties of taste and fragrance molecules, from the public databases Superscent, Flavornet, SuperSweet and BitterDB, taken collectively as flavours, in the perspective of the chemical space. We survey simple descriptor profiles in comparison with the public collections ChEMBL (bioactive small molecules), ZINC (commercial drug-like molecules) and GDB-13 (all possible organic molecules up to 13 atoms of C, N, O, S, Cl). A global analysis of the chemical space of flavours is also presented based on molecular quantum numbers (MQN) and SMILES fingerprints (SMIfp). While taste molecules span a very broad property range, fragrances occupy a narrow area of the chemical space consisting of generally very small and relatively nonpolar molecules distinct of standard drug molecules. Proximity searching in the chemical space is exemplified as a simple method to facilitate the search for new fragrances.