Generalized likelihood ratio (GLR) control charts are investigated for two types of statistical process monitoring (SPC) problems. The first part of this dissertation considers the problem of monitoring a normally distributed process variable when a special cause may produce a time varying linear drift in the mean. The design and application of a GLR control chart for drift detection is investigated. The GLR drift chart does not require specification of any tuning parameters by the practitioner, and has the advantage that, at the time of the signal, estimates of both the change point and the drift rate are immediately available. An equation is provided to accurately approximate the control limit. The performance of the GLR drift chart is compared to other control charts such as a standard CUSUM chart and a CUSCORE chart designed for drift detection. We also compare the GLR chart designed for drift detection to the GLR chart designed for sustained shift detection since both of them require only a control limit to be specified. In terms of the expected time for detection and in terms of the bias and mean squared error of the change-point estimators, the GLR drift chart has better performance for a wide range of drift rates relative to the GLR shift chart when the out-of-control process is truly a linear drift. The second part of the dissertation considers the problem of monitoring a linear functional relationship between a response variable and one or more explanatory variables (a linear profile). The design and application of GLR control charts for this problem are investigated. The likelihood ratio test of the GLR chart is generalized over the regression coefficients, the variance of the error term, and the possible change-point. The performance of the GLR chart is compared to various existing control charts. We show that the overall performance of the GLR chart is much better than other options in detecting a wide range of shift sizes. The existing control charts designed for certain shifts that may be of particular interest have several chart parameters that need to be specified by the user, which makes the design of such control charts more difficult. The GLR chart is very simple to design, as it is invariant to the choice of design matrix and the values of in-control parameters. Therefore there is only one design parameter (the control limit) that needs to be specified. Especially, the GLR chart can be constructed based on the sample size of n=1 at each sampling point, whereas other charts cannot be applied. Another advantage of the GLR chart is its built-in diagnostic aids that provide estimates of both the change-point and the values of linear profile parameters. / Ph. D.

No description available.

Les estuaires font partie des écosystèmes les plus productifs de la planète. Ils constituent une zone d'importants échanges de matières et jouent à ce titre un rôle capital pour les cycles biogéochimiques. Du fait de leur position à l'interface entre domaines continental et marin, ils sont également un lieu de passage obligatoire pour de nombreuses espèces de poissons migrateurs ou encore de nourricerie et de refuge. Ces systèmes sont contraints par de nombreuses pressions de natures diverses dont certaines en lien avec les activités humaines locales (p. ex. pêcheries, transport maritime, dragages, pollutions accrues, etc.). Ces altérations viennent s'ajouter à la variabilité intrinsèque de ces systèmes naturellement complexes (variabilité spatio-temporelle de différents facteurs environnementaux ; c.-à-d. température, salinité, turbidité, etc.). Toutefois, une intensification de ces d'altérations, dont une partie est associée à la composante climatique des changements globaux, est observée depuis ces dernières décennies. Ces modifications avérées et leurs incidences sur les communautés biologiques constituent un enjeu en matière de gestion des écosystèmes estuariens. Cependant, par manque de chroniques de données continues suffisamment longues, peu d'études sur l'évolution à long terme de systèmes estuariens ont été réalisées à ce jour. Pour mes travaux, j'ai eu à ma disposition un ensemble de chroniques uniques - acquises depuis 35 ans - associées à plusieurs suivis écologiques de l'estuaire de la Gironde (notre site atelier). L'objectif général de cette thèse est de caractériser l'évolution globale du système dans son ensemble (physico-chimie et biologie) et d'identifier les forçages impliqués. L'originalité de ces travaux, par rapport aux études spécifiques déjà menées, consiste à analyser l'évolution de l'ensemble des descripteurs caractéristiques du fonctionnement d'un système modèle et de qualifier la part relative aux forçages globaux. L'étude des séries chronologiques, via une approche end-to-end basée sur des analyses multivariées, permet d'identifier deux changements abrupts d'état de l'écosystème de l'estuaire de la Gironde, associés à des fluctuations hydroclimatiques à large échelle (bassin Atlantique nord) mais aussi à une échelle plus locale (estuaire). Des hypothèses relatives aux possibles mécanismes impliqués, reliant les altérations climatiques à large échelle aux altérations observées à l'échelle des communautés biologiques estuariennes, sont discutées. Un deuxième volet de la thèse, au travers d'une approche modélisatrice basée sur le concept de niche d'Hutchinson, permet d'identifier un changement de répartition spatiale, longitudinale, des populations de copépodes estuariennes en réponse au réchauffement des masses d'eau et à la marinisation de l'estuaire. Enfin un troisième volet de mes travaux s'est consacré à l'étude spécifique d'une espèce de copépode invasive : Acartia tonsa. Suivant une double approche modélisatrice permettant d'estimer à la fois les niches fondamentale et réalisée de l'espèce, ces travaux illustrent l'importance relative des changements climatiques qui ont facilité la colonisation et l'essor de cette espèce dans l'estuaire.

No description available.

Variability in the tropical ocean-atmospheric system causes global scale climate anomalies, most evident in the El Niño-Southern Oscillation’s coupled climate feedbacks. Despite being an area of high interest, many questions still remain regarding the west Pacific warm pool’s response to external forcing, particularly its response to increases in anthropogenic greenhouse gases. Paleoclimate reconstructions coupled with model simulations provide insight into the tropical Pacific’s role in past climate variability necessary to the development of robust climate projections. Most paleoclimate records, however, still lack the resolution, length, and chronological control to resolve rapid variability against a background of orbital-scale variations. Here we present stalagmite oxygen isotope (δ18O) reconstructions from Gunung Mulu National Park (4oN, 115oE ), in northern Borneo, that provide reproducible centennial-scale records of western Pacific hydrologic variability that are precisely U/Th-dated and continuous throughout most of the late Pleistocene (0-160 thousand years ago, kybp). The record comprises an entire glacial-interglacial cycle, which allows us to investigate orbital-scale climate forcings and compare two well-dated glacial terminations in the western tropical Pacific. The ice- volume-corrected δ18O records suggest that glacial boundary condtions, which include significantly lower atmospheric carbon dioxide levels, did not drive significant changes in Mulu rainfall δ18O. Similarly, Borneo stalagmite δ18O is poorly correlated to either global sea level shifts or Sunda Shelf areal exposure is not evident. The Borneo record does vary in phase with local mid-fall equatorial insolation, suggesting that precessional forcing may impart a strong influence on hydroclimate variability in the warm pool. This is best illustrated across Glacial Termination II, when the oscillation of equatorial fall insolation is large and out of phase with ice sheet decay. We also use a subset of well-dated, high-resolution stalagmite δ18O records from Mulu to investigate millennial-scale climate variability during Marine Isotope Stages 3-5 (30-100kybp). We find that regional convection likely decreased during the six massive iceberg discharges defined in the North Atlantic sediment records (“Heinrich events”). The inferred drying (increased stalagmite δ18O) during Heinrich events is consistent with a southward shift of the Intertropical Convergence Zone – the dominant paradigm to explain global climate anomalies originating in the north Atlantic (ref). However, any hydrologic variability related to Dansgaad-Oeschgar (D/O) events, millennial-scale sawtooth temperature anomalies of the last glacial period first evident in the Greenland ice records, is notably absent in the stalagmite records. . The Mulu stalagmite record’s absence of D/O signal, however, is in marked contrast to the regional west Pacific marine records and suggests D/O events and Heinrich events may be characterized by fundamentally different climate mechanisms and feedbacks.

Les industries lithiques de l’Holocène moyen, entre 8.000 et 4.000 BP dans le centre et le nord-est du Brésil ont souvent été décrites non pas en fonction des caractères présents mais de ceux absents : absence de lesma (pièce façonnée unifacialement), absence de céramique. Les traits techniques spécifiques à la période sont difficiles à cerner à cause de la diversité des artefacts. Décrites comme peu élaborées ou simples, ces industries interrogent pourtant par leurs caractères extrêmement différents de ceux de la période antérieure. Des hypothèses liées à des facteurs externes ont été évoquées pour expliquer de tels changements dans la culture matérielle : changement climatique avec une forte modification de l’environnement, remplacement de la population par une nouvelle vague de peuplement. Toutefois, il est difficile d’évaluer les modalités de cette rupture tant les analyses sur des assemblages de la période sont peu nombreux. Nous proposons dans ce travail de caractériser les systèmes techniques de l’Holocène moyen et par ce biais aborder le phénomène de rupture technique avec le technocomplexe Itaparica. Deux assemblages lithiques de l’Holocène moyen ont été étudiés selon une approche techno-fonctionnelle : la Toca Nova do Inharé dans la Serra da Capivara (Piauí) et le site Cajueiro (BA-RC-19) au nord-est du plateau central (Bahia).Les résultats obtenus mettent en évidence l’existence de caractères partagés et de variabilités importantes. L’organisation des différentes parties fonctionnelles des outils, structurée par la présence récurrente de pans abrupts, constitue un caractère commun fort. Les supports d’outils sont en revanche très variables selon les sites, expliquant cette impression d’absence d’outils clairement définis déjà soulignée.La confrontation de nos résultats aux données disponibles dans la littérature nous a permis d’affirmer ces caractères comme une spécificité des industries de l’Holocène moyen et de confirmer également l’existence d’une rupture technique importante avec le technocomplexe Itaparica. / During middle Holocene, between 8000 and 4000 BP, central and northern Brazil lithic industries were often described in terms of lacks more than tangible characteristics: lacks of “lesma” (unifacial shaping tool), lacks of ceramics. Specific technical features linked to that period are not easy to pinpoint because of a high level of artifacts diversity. Described as somewhat elaborate or simple, the specific nature of these industries highlights differences between this period and the previous one. To explain such a change in material culture, external factor hypothesis were proposed: huge environment modification due to climate change, new population that replaces another one. However, it is uneasy to evaluate break modalities regarding few studies on these period assemblages.Our work proposes to characterize the technical systems for the middle Holocene, and, by this way, to broach the technical break phenomenon using Itaparica techno-complex. Two middle Holocene lithic assemblages were study in a techno-functional way: la Toca Nova do Inharé, in Serra da Capivara (Piauí) and the Cajueiro site (BA-RC-19) northeastern of the central plateau (Bahia). Results reveal shared characteristics and significant variability. Organization of different functional parts of tools outlined around recurrent presence of abrupt sides, form a strong shared feature. Tool blanks have, however, high degree of variability according to different sites, explaining the lack of strictly defined tools. Our results compared to available literature data allow us to confirm that these features are specific to these middle Holocene industries and also to bear out the existence of a significant technical break with the Itaparica techno-complex.

Les technologies multimédias ont récemment connues une grande évolution surtout avec la croissance rapide d'internet ainsi que la création quotidienne de grands volumes de données vidéos. Tout ceci nécessite de nouvelles méthodes performantes permettant d'indexer, de naviguer, de rechercher et de consulter les informations stockées dans de grandes bases de données multimédia. La récupération de données basée sur le contenu vidéo, qui est devenue un domaine de recherche très actif durant cette décennie, regroupe les différentes techniques conçues pour le traitement de la vidéo. Dans le cadre de cette thèse de doctorat, nous présentons des applications permettant la segmentation temporelle d'une vidéo ainsi que la récupération d'information pertinente dans une séquence vidéo. Une fois le processus de classification effectué, il devient possible de rechercher l'information utile en ajoutant de nouveaux critères, et aussi de visualiser l'information d'une manière appropriée permettant d'optimiser le temps et la mémoire. Dans une séquence vidéo, le plan est considéré comme l'unité élémentaire de la vidéo. Un plan est défini comme une suite d'image capturée par une même caméra représentant une action dans le temps. Pour composer une vidéo, plusieurs plans sont regroupés en utilisant des séquences de transitions. Ces transitions se catégorisent en transitions brusques et transitions progressives. Détecter les transitions présentes dans une séquence vidéo a fait l'objet de nos premières recherches. Plusieurs techniques, basées sur différents modèles mathématiques, ont été élaborées pour la détection des changements de plans. L'utilisation de la décomposition en valeur singulière (SVD) ains que la norme Frobenius ont permis d'obtenir des résultats précis en un temps de calcul réduit. Le résumé automatique des séquences vidéo est actuellement un sujet d'une très grande actualité. Comme son nom l'indique, il s'agit d'une version courte de la vidéo qui doit contenir l'essentiel de l'information, tout en étant le plus concis possible. Ils existent deux grandes familles de résumé : le résumé statique et le résumé dynamique. Sélectionner une image représentative de chaque plan permet de créer un scénarimage. Ceci est considéré comme étant un résumé statique et local. Dans notre travail, une méthode de résumé globale est proposée. / With the recent advancement in multimedia technologies, in conjunction with the rapid increase of the volume of digital video data and the growth of internet ; it has becom mandatory to have the hability browse and search through information stored in large multimedia databases. For this purpose, content based video retrieval (CBVR) has become an active area of research durinf the last decade. The objective of this thesis is to present applications for temporal video segmentation and video retrieval based on different mathematical models. A shot is considered as the elementary unit of a video, and is defined as a continuous sequence of frames taken from a single camera, representing an action during time. The different types of transitions that may occur in a video sequence are categorized into : abrupt and gradual transition. In this work, through statistical analysis, we segment a video into its constituent units. This is achieved by identifying transitions between adjacent shots. The first proposed algorithm aims to detect abrupt shot transitions only by measuring the similarity between consecutive frames. Given the size of the vector containing distances, it can be modeled by a log normal distribution since all the values are positive. Gradual shot transition identification is a more difficult task when compared to cut detection. Generally, a gradual transition may share similar characteristics as a dynamic segment with camera or object motion. In this work, singular value decomposition (SVD) is performed to project features from the spatial domain to the singular space. Resulting features are reduced and more refined, which makes the remaining tasks easier. The proposed system, designed for detecting both abrupt and gradual transitions, has lead to reliable performances achieving high detection rates. In addition, the acceptable computational time allows to process in real time. Once a video is partitioned into its elementary units, high-level applications can be processed, such as the key-frame extraction. Selecting representative frames from each shot to form a storyboard is considered as a static and local video summarization. In our research, we opted for a global method based on local extraction. Using refined centrist features from the singular space, we select representative frames using modified k-means clustering based on important scenes. This leads to catch pertinent frames without redoudancy in the final storyboard.

Peat from the temperate regions has been used for paleoenvironmental reconstruction using diverse proxies for over a century now. Peat is rare and severely understudied in the tropics. The montane peat bogs of the Nilgiris, southern India have been found to preserve global climatic signals including the Holocene Optimum and the Last Glacial Maximum (LGM). At Sandynallah, one of the oldest peat accumulations in the world at >40 kyr BP, we had undertaken a high resolution paleoenvironmental reconstruction using multiple proxies which are yet to be evaluated in the tropical context. The study consists of 3 main objectives, i. Establishing an accurate high resolution chronology for the peat profile using radiocarbon dating, ii. Extracting vegetation and climate information from C/N ratio and Rock-Eval indices and iii. Using elemental profiles to establish the utility of inorganic geochemical proxies for processes such as weathering and dust transport. High resolution chronology for the site was built using Accelerator Mass Spectrometry (AMS) radiocarbon dates. To improve accuracy of the age-depth model, we also tested 4 samples for the effects of the AAA (Acid-Alkali-Acid) extraction method, the most common pre-treatment method for peat in the world. We compared ages on bulk samples (acid washed) and AAA treated samples from 4 different depths. We find that for all depths, the ages of the untreated samples do not lie within the internal uncertainty window and differ from the AAA treated sample age by at least an order of magnitude of the internal error, if not more. Based on these results we argue that the internal error should be used in conjunction with a reliable estimate of external error in an age-depth model for more realistic dating of paleoclimatic events. C/N ratios were explored for their paleoclimatic potential in conjunction with Rock-Eval indices and it was found that decomposition in tropical peat, as opposed to temperate peat, may not be sensitive to climatic perturbations. Inorganic geochemical proxies were also evaluated through this study. We see that the major and trace elements, except the lanthanide series do not show many significant trends for paleoenvironmental interpretation. But the lanthanides show some promise for identifying potential sources of dust and weathered material. Our study has addressed the gap in knowledge about the utility of recent geochemical proxies in tropical peat and has attempted to provide a solution to improve reliability in constructing age-depth models.

Este trabajo de investigación aporta evidencia empírica sobre el impacto de cambios en la política monetaria de la Reserva Federal de los Estados Unidos en la cantidad de entrada de capitales en economías emergentes. Utilizando una muestra de países en desarrollo entre los años 2000 al 2019, se logra establecer que los influjos de capitales de las economías emergentes se reducen considerablemente en aquellos periodos donde hubo un recorte de la expansión monetaria por parte de la Reserva Federal de EE. UU aumentando la probabilidad de ocurrencia de episodios de reversión abrupta de entrada de capitales. Este hallazgo empírico resulta de gran relevancia debido al actual contexto económico donde variaciones en la política monetaria de la Reserva Federal es altamente probable, además de la importancia que tienen los influjos de capitales en el crecimiento económico. Asimismo, a partir de estos resultados, se establece una serie de recomendaciones de política económica. / This research work provides empirical evidence on the impact of changes in the monetary policy of the United States Federal Reserve on the amount of capital inflows in emerging economies. Using a sample of developing countries between the years 2000 and 2019, it is possible to establish that capital inflows from emerging economies are considerably reduced in those periods where there was a cut in monetary expansion by the US Federal Reserve. increasing the probability of occurrence of episodes of abrupt reversal of capital inflows. This empirical finding is highly relevant due to the current economic context where variations in the monetary policy of the Federal Reserve are highly probable, in addition to the importance of capital inflows in economic growth. Likewise, based on these results, a series of economic policy recommendations are established. / Tesis

When monitoring a binary process proportion p, it is usually assumed that the binary observations are independent. However, it is very common that the observations are correlated with p being the correlation between two successive observations. The first part of this research investigates the problem of monitoring p when the binary observations follow a first-order two-state Markov chain model with p remaining unchanged. A Markov Binary GLR (MBGLR) chart with an upper bound on the estimate of p is proposed to monitor a continuous stream of autocorrelated binary observations treating each observation as a sample of size n=1. The MBGLR chart with a large upper bound has good overall performance over a wide range of shifts. The MBGLR chart is optimized using the extra number of defectives (END) over a range of upper bounds for the MLE of p. The numerical results show that the optimized MBGLR chart has a smaller END than the optimized Markov binary CUSUM. The second part of this research develops a CUSUM-pp chart and a GLR-pp chart to monitor p and p simultaneously. The CUSUM-pp with two tuning parameters is designed to detect shifts in p and p when the shifted values are known. We apply two CUSUM-pp charts as a chart combination to detect increases in p and increases or decreases in p. The GLR-pp chart with an upper bound on the estimate of p, and an upper bound and a lower bound on the estimate of p works well when the shifts are unknown. We find that the GLR-pp chart has better overall performance. The last part of this research investigates the problem of monitoring p with p remains at the target value when the correlated binary observations are aggregated into samples with n>1. We assume that samples are independent and there is correlation between the observations in a sample. We proposed some GLR and CUSUM charts to monitor p and the performance of the charts are compared. The simulation results show MBNGLR has overall better performance than the other charts. / Ph. D.