Dyer Island is thought to host one of the most abundant populations of white sharks on the planet; this is often credited to the large (55 – 60,000) Cape fur seal colony at Geyser Rock. Yet relatively little work has ever been produced from the area. This may be attributed to the harshness in its location as a study site, exposed to wind and swell from west to east which limits research periods. This study accounts for over 220 hrs of manual tracking at Dyer Island with a further 68 added from the inshore shallow areas of the bay. Sharks focused their movements and habitat use to reefs or channels that allowed access to Cape fur seals. Movement- Based Kernel Estimates (MKDE) were used to compute home range estimates for shark movements through and around the heterogeneous structures of Dyer Island and Geyser Rock. Inshore two core areas were revealed, one being the major reef system at Joubertsdam and the other at a kelp reef where the tracked shark had fed on a Cape fur seal. At Dyer Island one core area was identified in a narrow channel, ‘Shark Alley’, here a second tracked shark foraged for entire days within meters of rafting Cape fur seals. Rate of Movement (ROM) and Linearity (LI) of tracks were low during daytime and movements were focused around areas such as Shark Alley or other areas close to the seal colony before moving into deeper water or distant reefs with higher rates of ROM and LI at night. If moonlight was strong foraging would take place to the south of Geyser Rock but with higher ROM and LI than observed during the day. Foraging patterns in this study contrast studies from other sites in South Africa and home range and activity areas were comparatively much smaller than observed in Mossel Bay. It has been established that several known white sharks forage at Dyer Island and the other studied aggregation sites, such differences in foraging would suggest that they are able to adapt their foraging behaviour to suit the environment they are in; making them site specific in their foraging ecology. Both satellite and acoustic telemetry are revealing aggregation hotspots of white sharks in South Africa. It is important that such information is used to assist the recovery of the species which has been protected since 1991, yet is rarely considered in planning of coastal developments. / Dissertation (MSc)--University of Pretoria, 2012. / Zoology and Entomology / MSc / Unrestricted

The use of past archaeological survey data for examination of landscape dynamics became very popular during the last decades of the 20th century, when Geographical Information Systems analysis were introduced in archaeology. In the present thesis, past survey data from Northern Messenia’s Middle and Late Helladic periods are combined to the topography and the environment of the region. These data are examined under the GIS prism, which attempts to update our knowledge on this geographical area.             The main scope of this thesis is to examine potential settlement patterns and land use, connectivity between sites and sites’ hierarchies. In the first query, the Kernel Density analysis has been used for estimating settlements’ patterns, and to consequently estimate preference of specific topographical features for land use, such as slope. Based on the patterns formed as a result of the analysis conducted in the first query, connectivity and hierarchy between sites is being tested with the use of cost connectivity and visibility tools.             The outcome of this analysis shows that the inhabitants of the past were significantly interacting with the landscape, since they preferred to nest around the protective slopes of the Soulima and the Kyparissian valleys. The area around their settlements reveals that these inhabitants opted to cultivate in flat or marginal land, while visibility from the sites seems to be an important factor for monitoring the region. However, it has been proved that a good number of collaborating sites were required to supervise the entire territory, which disproves any hierarchical ranking between them. On the other hand, connectivity depicts potential movement over Northern Messenia’s terrain and indicates that a few sites in the heart of the study area were to be considered as panoptic meeting grounds of the eastern and the western side. In conclusion, the overall analysis reveals a potential spatial bond between sites rather than a relationship based on rivalry.

Nous considérons le problème d’estimation de densités conditionnelles en modérément grandes dimensions. Beaucoup plus informatives que les fonctions de régression, les densités condi- tionnelles sont d’un intérêt majeur dans les méthodes récentes, notamment dans le cadre bayésien (étude de la distribution postérieure, recherche de ses modes...). Après avoir rappelé les problèmes liés à l’estimation en grande dimension dans l’introduction, les deux chapitres suivants développent deux méthodes qui s’attaquent au fléau de la dimension en demandant : d’être efficace computation- nellement grâce à une procédure itérative gloutonne, de détecter les variables pertinentes sous une hypothèse de parcimonie, et converger à vitesse minimax quasi-optimale. Plus précisément, les deux méthodes considèrent des estimateurs à noyau bien adaptés à l’estimation de densités conditionnelles et sélectionnent une fenêtre multivariée ponctuelle en revisitant l’algorithme glouton RODEO (Re- gularisation Of Derivative Expectation Operator). La première méthode ayant des problèmes d’ini- tialisation et des facteurs logarithmiques supplémentaires dans la vitesse de convergence, la seconde méthode résout ces problèmes, tout en ajoutant l’adaptation à la régularité. Dans l’avant-dernier cha- pitre, on traite de la calibration et des performances numériques de ces deux procédures, avant de donner quelques commentaires et perspectives dans le dernier chapitre. / We consider the problem of conditional density estimation in moderately large dimen- sions. Much more informative than regression functions, conditional densities are of main interest in recent methods, particularly in the Bayesian framework (studying the posterior distribution, find- ing its modes...). After recalling the estimation issues in high dimension in the introduction, the two following chapters develop on two methods which address the issues of the curse of dimensionality: being computationally efficient by a greedy iterative procedure, detecting under some suitably defined sparsity conditions the relevant variables, while converging at a quasi-optimal minimax rate. More precisely, the two methods consider kernel estimators well-adapted for conditional density estimation and select a pointwise multivariate bandwidth by revisiting the greedy algorithm RODEO (Regular- isation Of Derivative Expectation Operator). The first method having some initialization problems and extra logarithmic factors in its convergence rate, the second method solves these problems, while adding adaptation to the smoothness. In the penultimate chapter, we discuss the calibration and nu- merical performance of these two procedures, before giving some comments and perspectives in the last chapter.

Les mesures de vibrations sont l’une des données les plus pertinentes pour détecter des anomalies sur les moteurs. Les vibrations sont acquises sur banc d’essai en phase d’accélération et de décélération pour assurer la fiabilité du moteur à la sortie de la chaine de production. Ces données temporelles sont converties en spectrogrammes pour permettre aux experts d’effectuer une analyse visuelle de ces données et de détecter les différentes signatures atypiques. Les sources vibratoires correspondent à des raies sur les spectrogrammes. Dans cette thèse, nous avons mis en place un outil d’aide à la décision automatique pour analyser les spectrogrammes et détecter tout type de signatures atypiques, ces signatures ne proviennent pas nécessairement d’un endommagement du moteur. En premier lieu, nous avons construit une base de données numérique de spectrogrammes annotés. Il est important de noter que les signatures inusuelles sont variables en forme, intensité et position et se trouvent dans un faible nombre de données. Par conséquent, pour détecter ces signatures, nous caractérisons les comportements normaux des spectrogrammes, de manière analogue aux méthodes de détection de nouveautés, en représentant les patchs des spectrogrammes sur des dictionnaires comme les curvelets et la Non-negative matrix factorization (NMF), ainsi qu’en estimant la distribution de chaque point du spectrogramme à partir de données normales dépendamment ou non de leur voisinage. La détection des points atypiques est réalisée par comparaison des données tests au modèle de normalité estimé sur des données d’apprentissage normales. La détection des points atypiques permet la détection des signatures inusuelles composées par ces points. / Vibration measurements are one of the most relevant data for detecting anomalies in engines. Vibrations are recorded on a test bench during acceleration and deceleration phases to ensure the reliability of every flight engine at the end of the production line. These temporal signals are converted into spectrograms for experts to perform visual analysis of these data and detect any unusual signature. Vibratory signatures correspond to lines on the spectrograms. In this thesis, we have developed a decision support system to automatically analyze these spectrograms and detect any type of unusual signatures, these signatures are not necessarily originated from a damage in the engine. Firstly, we have built a numerical spectrograms database with annotated zones, it is important to note that data containing these unusual signatures are sparse and that these signatures are quite variable in shape, intensity and position. Consequently, to detect them, like in the novelty detection process, we characterize the normal behavior of the spectrograms by representing patches of the spectrograms in dictionaries such as the curvelets and the Non-negative matrix factorization (NMF) and by estimating the distribution of every points of the spectrograms with normal data depending or not of the neighborhood. The detection of the unusual points is performed by comparing test data to the model of normality estimated on learning normal data. The detection of the unusual points allows the detection of the unusual signatures composed by these points.

Tropical cyclone (TC)-spawned tornadoes in Florida were analyzed to determine patterns of occurrence based on storm and geographic features. Tornadoes were determined to be associated with a landfalling or nearby TC if a tornado occurred within 800 km of the TC’s center of circulation. TC-tornadoes were analyzed for patterns based on distance and angle from TC’s center, topographic influences, population biases, and influence based on time of landfall. Most TC-Tornadoes tend to occur more often before landfall than after. It was discovered that tornadoes have occurred in different areas with respect to the bearing from the center depending on the landfall location and time of landfall. It was also discovered that land use type, and elevation had little to do with TC-Tornado occurrence. The results do suggest some population bias. The findings will be a guide for operational meteorologists to aid in forecasting likely tornadogenesis from TCs.

Yes / This paper investigates the interaction phenomena of the coupled axons while the mutual coupling factor is presented as a pairwise description. Based on the Hodgkin-Huxley model and the coupling factor matrix, the membrane potentials of the coupled myelinated/unmyelinated axons are quantified which implies that the neural coupling can be characterised by the presented coupling factor. Meanwhile the equivalent electric circuit is supplied to illustrate the physical meaning of this extended model. In order to estimate the coupling factor, a data-based iterative learning identification algorithm is presented where the Rényi entropy of the estimation error has been minimised. The convergence of the presented algorithm is analysed and the learning rate is designed. To verified the presented model and the algorithm, the numerical simulation results indicate the correctness and the effectiveness. Furthermore, the statistical description of the neural coupling, the approximation using ordinary differential equation, the measurement and the conduction of the nerve signals are discussed respectively as advanced topics. The novelties can be summarised as follows: 1) the Hodgkin-Huxley model has been extended considering the mutual interaction between the neural axon membranes, 2) the iterative learning approach has been developed for factor identification using entropy criterion, and 3) the theoretical framework has been established for this class of system identification problems with convergence analysis. / This work was supported in part by the National Natural Science Foundation of China (NSFC) under Grant 51807010, and in part by the Natural Science Foundation of Hunan under Grant 1541 and Grant 1734. / Research Development Fund Publication Prize Award winner, Nov 2020.

SPECT (Single Photon Emission Computed Tomography) is the modality of choice for myocardial perfusion imaging due to the high sensitivity and specificity, and the lower cost of equipment and radiotracers compared to PET. Dynamic SPECT imaging provides new possibilities for myocardial perfusion imaging by encoding more information in the reconstructed images in the form of time-activity functions. The recent introduction of small solid-state SPECT cameras using multiple pinhole collimators, such as the GE Discovery NM 530c, offers the ability to obtain accurate myocardial perfusion information with markedly decreased acquisition times and offers the possibility to obtain quantitative dynamic perfusion information. This research targets two aspects of dynamic SPECT imaging with the intent of contributing to the improvement of projection and reconstruction methods. First, we propose an adaptation of distance-driven projection to SPECT imaging systems using single-pinhole collimator detectors. The proposed distance-driven projection approach accounts for the finite size of the pinhole, the possibly coarse discretization of the detector and object spaces, and the tilt of the detector surface. We evaluate the projection method in terms of resolution and signal to noise ratio (SNR). We also propose two maximum a posteriori (MAP) iterative image reconstruction methods employing kernel density estimators. The proposed reconstruction methods cluster time-activity functions (or intensity values) by their spatial proximity and similarity, each of which is determined by spatial and range scaling parameters respectively. The results of our experiments support our belief that the proposed reconstruction methods are especially effective when performing reconstructions from low-count measurements. / Thesis / Doctor of Philosophy (PhD)

No description available.

This thesis investigates some problems related to the form and shape of statistical distributions with the main focus on goodness of fit and bump hunting. A bump is a distinctive characteristic of distributional shape. A search for bumps, or bump hunting, in a probability density function (PDF) has long been an important topic in statistical research. We introduce a new definition of a bump which relies on the notion of the curvature of a planar curve. We then propose a new method for bump hunting which is based on a kernel density estimator of the unknown PDF. The method gives not only the number of bumps but also the location of their centers and base points. In quantitative risk applications, the selection of distributions that properly capture upper tail behavior is essential for accurate modeling. We study tests of distributional form, or goodness-of-fit (GoF) tests, that assess simple hypotheses, i.e., when the parameters of the hypothesized distribution are completely specified. From theoretical and practical perspectives, we analyze the limiting properties of a family of weighted Cramér-von Mises GoF statistics W2 with weight function psi(t)=1/(1-t)^beta (for beta<=2) which focus on the upper tail. We demonstrate that W2 has no limiting distribution. For this reason, we provide a normalization of W2 that leads to a non-degenerate limiting distribution. Further, we study W2 for composite hypotheses, i.e., when distributional parameters must be estimated from a sample at hand. When the hypothesized distribution is heavy-tailed, we examine the finite sample properties of W2 under the Chen-Balakrishnan transformation that reduces the original GoF test (the direct test) to a test for normality (the indirect test). In particular, we compare the statistical level and power of the pairs of direct and indirect tests. We observe that decisions made by the direct and indirect tests agree well, and in many cases they become independent as sample size grows. / Thesis / Doctor of Philosophy (PhD)

The endangered northern long-eared bat (<i>Myotis septentrionalis</i>) has experienced severe population declines since the 2007 introduction of <i>Pseudogymnoascus destructans</i>, the fungal pathogen that causes white-nose syndrome (WNS). Due to continued mortality, failed recruitment, and range fragmentation, effective species conservation would benefit from local and regional research related to habitat selection and suitability, seasonal activity, and large-scale estimates of continued occupancy across the post-WNS landscape. In this dissertation, I explored maternity colony home range size and habitat selection, both coastal and interior seasonal activity patterns, habitat suitability, and species occupancy trends in the Mid-Atlantic and Northeastern United States. Research from the Coastal Plain of North Carolina, based on diurnal roost locations, revealed that core and peripheral home range estimates were large but comparable to areas of contiguous forest cover in other regions of the species' range. Prior to juvenile volancy, female northern long-eared bats selected red maple (<i>Acer rubrum</i>), water ash (<i>Fraxinus caroliniana</i>), and loblolly pine (<i>Pinus taeda</i>) as day-roosts, but then used sweetgum (<i>Liquidambar styraciflua</i>), swamp bay (<i>Persea palustris</i>), and water tupelo (<i>Nyssa aquatica</i>) after juvenile volancy. At the second-order spatial scale (i.e., home range within a region), roosting home ranges were associated with woody wetlands farther from anthropogenic development and open water. However, within the third-order scale (i.e., core home range within a periphery), northern long-eared bats selected woody wetlands adjacent to intact upland forests, areas containing shorter trees and occurring proximal to open water. Research utilizing passive acoustic monitoring on the Coastal Plain of both Virginia and North Carolina found that northern long-eared bat relative activity was highest in areas containing greater proportions of woody wetlands and upland pine-dominated evergreen forests. Conversely, the likelihood to record northern long-eared bats was associated with smaller proportions of woody wetlands and open water resources, emphasizing the importance of upland forests adjacent to these features. I also observed a higher probability of recording northern long-eared bats during non-winter seasons and when temperatures were between 10 and 25 °C, potentially highlighting an optimal thermoneutral zone for the species regionally. Research using presence data from the Monongahela National Forest (MNF) in the Central Appalachians of West Virginia found that northern long-eared bat habitat, whether occupied or not, was abundant (43.1% of the MNF) and widely dispersed. Northern long-eared bat habitat on the MNF was often characterized as mature, interior mixed mesophytic forests. Research using passive acoustic surveys associated with hibernacula in western Virginia found that northern long-eared bats were most active near hibernacula during warmer weeks of the fall swarm and spring emergence, when rainfall was low. Similarly, the probability of northern long-eared bat activity was highest near hibernacula during the spring/summer season. However, unlike relative activity, the likelihood of recording northern long-eared bats was associated with more heterogeneous, interior forests. Finally, research using a combination of long-term repeated and single-season clearance survey data from New England found that northern long-eared bat occupancy was highest on steep hillsides in contiguous forested landscapes with minimal anthropogenic development. My results also indicated higher occupancy of northern long-eared bats in warmer regions, particularly along the New England coastline and offshore islands. These findings collectively stress the importance of managing areas of large core forest to promote sustainable roost formation and productive foraging areas, often associated with dynamic ecotones, to support the survival and recovery of northern long-eared bats in the post-WNS Mid-Atlantic and Northeast. / Doctor of Philosophy / The northern long-eared bat was once one of the most common bat species in North America. However, due to population impacts caused by white-nose syndrome (WNS) the species is now recognized as endangered by the United States Fish and Wildlife Service. In this dissertation, I explore habitat selection and suitability, seasonal activity, and occupancy trends of remnant populations of northern long-eared bats in the Mid-Atlantic and Northeast of the United States, representing approximately 25% of the species' historical range. Specifically, I examined home range size and habitat selection of a reproductively successful maternity colony on the Mid-Atlantic Coastal Plain, overwintering activity at both interior and coastal sites, habitat suitability in the Central Appalachians, and spatial occupancy trends across New England. My research found that reproductive northern long-eared bats actively selected for specific cover features and displayed notable shifts in roost tree selection throughout the maternity season on the Coastal Plain of North Carolina. My research also indicated that northern long-eared bats were more active during non-winter seasons, favoring moderate temperatures and areas of forested wetlands adjacent to evergreen forests in coastal North Carolina and Virginia. In Virginia, northern long-eared bat activity near mountain hibernacula was highest in areas of greater landscape richness prior to and after hibernation. Finally, my research from New England found that northern long-eared bat occupancy was highest in steeper, forested landscapes in the absence of human development. Overall, this research highlights the need to conserve and manage forest ecosystems to promote recovery of the endangered northern long-eared bat. Conservation and management efforts informed by population status, activity trends, and habitat associations will be invaluable in guiding species recovery efforts.