Дисертації з теми "Prédiction temporelle"
Оформте джерело за APA, MLA, Chicago, Harvard та іншими стилями
Ознайомтеся з топ-50 дисертацій для дослідження на тему "Prédiction temporelle".
Біля кожної праці в переліку літератури доступна кнопка «Додати до бібліографії». Скористайтеся нею – і ми автоматично оформимо бібліографічне посилання на обрану працю в потрібному вам стилі цитування: APA, MLA, «Гарвард», «Чикаго», «Ванкувер» тощо.
Також ви можете завантажити повний текст наукової публікації у форматі «.pdf» та прочитати онлайн анотацію до роботи, якщо відповідні параметри наявні в метаданих.
Переглядайте дисертації для різних дисциплін та оформлюйте правильно вашу бібліографію.
Hadjoudja, Abdelkader. "Macrogénération et prédiction temporelle sur les réseaux programmables CPLD." Grenoble INPG, 1997. http://www.theses.fr/1997INPG0177.
Повний текст джерелаDeregnaucourt, Thomas. "Prédiction spatio-temporelle de surfaces issues de l'imagerie en utilisant des processus stochastiques." Thesis, Université Clermont Auvergne (2017-2020), 2019. http://www.theses.fr/2019CLFAC088.
Повний текст джерелаThe prediction of a surface is now an important problem due to its use in multiple domains, such as computer vision, the simulation of avatars for cinematography or video games, etc. Since a surface can be static or dynamic, i.e. evolving with time, this problem can be separated in two classes: a spatial prediction problem and a spatio-temporal one. In order to propose a new approach for each of these problems, this thesis works have been separated in two parts.First of all, we have searched to predict a static surface, which is supposed cylindrical, knowing it partially from curves. The proposed approach consisted in deforming a cylinder on the known curves in order to reconstruct the surface of interest. First, a correspondence between known curves and the cylinder is generated with the help of shape analysis tools. Once this step done, an interpolation of the deformation field, which is supposed Gaussian, have been estimated using maximum likelihood and Bayesian inference. This methodology has then been applied to real data from two domains of imaging: medical imaging and infography. The obtained results show that the proposed approach exceeds the existing methods in the literature, with better results using Bayesian inference.In a second hand, we have been interested in the spatio-temporal prediction of dynamic surfaces. The objective was to predict a dynamic surface based on its initial surface. Since the prediction needs to learn on known observations, we first have developed a spatio-temporal surface analysis tool. This analysis is based on shape analysis tools, and allows a better learning. Once this preliminary step done, we have estimated the temporal deformation of the dynamic surface of interest. More precisely, an adaptation, with is usable on the space of surfaces, of usual statistical estimators has been used. Using this estimated deformation on the initial surface, an estimation of the dynamic surface has been created. This process has then been applied for predicting 4D expressions of faces, which allow us to generate visually convincing expressions
Langlois, Sébastien. "Prédiction des vibrations éoliennes d'un système conducteur-amortisseur avec une méthode temporelle non linéaire." Thèse, Université de Sherbrooke, 2013. http://hdl.handle.net/11143/6133.
Повний текст джерелаArrouet, Alana. "Exploration de la prédiction temporelle associée à la motricité chez les individus neurotypiques et neuro-atypiques." Electronic Thesis or Diss., Strasbourg, 2024. http://www.theses.fr/2024STRAJ067.
Повний текст джерелаThe aim of this thesis was to explore the impact of temporal prediction on movement planning and execution. We used motor tasks in which participants stopped an index finger movement in response to a predictable target signal and examined how this prediction influenced both movement preparation and stopping execution. In neurotypical individuals, our findings revealed multiple temporal prediction mechanisms operating simultaneously: one linked to motor commands affecting preparation, a sensorimotor prediction influencing execution, and an independent prediction reflecting cognitive anticipation. Sensorimotor temporal prediction evolves with development and appears to be impaired in individuals at high genetic risk of psychotic conversion. In people with schizophrenia, preliminary findings suggest that performing a movement may help restore temporal prediction abilities. This thesis provides insights into how temporal predictions are integrated into motor programs and raises questions about the mechanisms underlying sensorimotor integration
Lajugie, Rémi. "Prédiction structurée pour l’analyse de données séquentielles." Thesis, Paris, Ecole normale supérieure, 2015. http://www.theses.fr/2015ENSU0024/document.
Повний текст джерелаIn this manuscript, we consider structured machine learning problems and consider more precisely the ones involving sequential structure. In a first part, we consider the problem of similarity measure learning for two tasks where sequential structure is at stake: (i) the multivariate change-point detection and (ii) the time warping of pairs of time series. The methods generally used to solve these tasks rely on a similarity measure to compare timestamps. We propose to learn a similarity measure from fully labelled data, i.e., signals already segmented or pairs of signals for which the optimal time warping is known. Using standard structured prediction methods, we present algorithmically efficient ways for learning. We propose to use loss functions specifically designed for the tasks. We validate our approach on real-world data. In a second part, we focus on the problem of weak supervision, in which sequential data are not totally labeled. We focus on the problem of aligning an audio recording with its score. We consider the score as a symbolic representation giving: (i) a complete information about the order of events or notes played and (ii) an approximate idea about the expected shape of the alignment. We propose to learn a classifier for each note using this information. Our learning problem is based onthe optimization of a convex function that takes advantage of the weak supervision and of the sequential structure of data. Our approach is validated through experiments on the task of audio-to-score on real musical data
Jiga-Boy, Gabriela-Maria. "Adaptative thinking about the future : temporal construal, health-related behaviour and perceived temporal distance." Grenoble 2, 2008. http://www.theses.fr/2008GRE29032.
Повний текст джерелаThis thesis investigated a "functional paradox" between the importance we ascribe to being future-oriented in order to function efficiently and the quality of our future outlooks. The underlying processes signal that one is most of the time wrong when predicting future behaviours. We explored the way individuals construe future events and act on the basis of these representations (Chapter I). Drawing on construal level theory (Liberman & Trope, 1998), we first replicated the relationship between temporal distance and concrete/abstract construal of future events (Chapter II). Next, we failed to characterize construal of health-related behaviour using this pattern. These events were construed more concretely/abstractly irrespective of temporal distance of their enactment. We further investigated the factors shaping construal level and found that it varies with the personal relevance of information, the individual’s goals, and the goal-related actions he (she) is engaged in (Chapter III). Finally, in order to situate our actions in time, we explored how individuals perceive temporal distance to future events (Chapter IV). We found that effort to be invested in a future event shapes the perception of when the event happens: more effortful events are felt to be happening earlier than less effortful events. Overall, the findings reported in this thesis bring information about what underlies our future outlooks. They suggest that the way we represent our future actions is grounded in their personal context – in other words that our interests, motivations, and actions' personal relevancies shape the concreteness of our future actions’ construal
Hafid, Mohamed. "Prédiction par transfert inverse de l'évolution temporelle du front de solidification : applications aux réacteurs métallurgiques et à la cryochirurgie." Thèse, Université de Sherbrooke, 2017. http://hdl.handle.net/11143/10581.
Повний текст джерелаBossavy, Arthur. "Caractérisation et prédiction probabiliste des variations brusques et importantes de la production éolienne." Phd thesis, Ecole Nationale Supérieure des Mines de Paris, 2012. http://pastel.archives-ouvertes.fr/pastel-00803234.
Повний текст джерелаBehlouli, Hassan. "Apprentissages auto-améliorants et modélisation de la dynamique temporelle de données évolutives par réseaux de neurones : application au diagnostic et la prédiction en électrocardiologie quantitative." Lyon, INSA, 1998. http://www.theses.fr/1998ISAL0034.
Повний текст джерелаWe present various methodologies to improve decision making on follow-up patient data and their validation in the field of quantitative electrocardiology. First, we propose an extension to the classical Pattern Recognition supervised learning model by introducing a self-improving concept based on information min. Ing from undocumented datasets. Then we apply this concept to the particular case of neural networks based supervised learning and we propose a self-improving learning methodology integrating iteratively, in the initial learning set, undocumented data that are extracted from databases not validated by experts. This method involves different concepts such as neural network combination, rejection of ambiguous cases and control of the learning process by cross-validation. Using this approach for the categorisation of cardiac diseases we could significantly improve the performance of the original classifiers. Secondly, we developed a methodology based on neural networks to model the dynamic behavior of the heart particularly for predicting one of the main descriptors of the ventricular repolarisation, i. E. : the QT interval as a function of the RR interval that represents the inverse of heart rate. An initial evaluation on a series of sequences of 30 electrocardiograms (3D ECG) continuously '1 recorded over 24 hours allowed to demonstrate the pertinence of the models and to study the ray of some parameters (e. G. Memory effect and noise level) on the prediction quality of this model We conclude by presenting another outcome of our work, a series of generic analysis processing tool s that were integrated into the MATIS environment (Mathematical Tools Integration Software), which is a fundamental building black for the future workstation of the research cardiologist
Idir, Mohamed Yacine. "Analyse et développement de modèles statistiques pour l'estimation et la prédiction spatiale et temporelle de la pollution atmosphérique à partir de données issues de capteurs mobiles." Electronic Thesis or Diss., université Paris-Saclay, 2023. http://www.theses.fr/2023UPASG107.
Повний текст джерелаUrban air pollution, a global health crisis causing millions of deaths every year, makes accurate mapping of this phenomenon not only relevant, but vital to public health.Currently, air quality is measured by fixed air quality monitoring stations. These reference stations provide a highly accurate measure of air quality, at the cost of limited spatial coverage.The idea of using new low-cost sensors developed from recent technological advances, smaller in size and incorporating a global positioning system (GPS), quickly emerged. This gives scientists additional tools to refine spatio-temporal maps of air pollution and create new datasets providing information on air quality that was previously unavailable.Generating precise air quality maps using these low-cost sensors presents several major challenges. These challenges are mainly related to the nature of the phenomenon being studied, and to the accuracy and volume of the data.Given these difficulties, it is important to know how to combine all these fuzzy data sources to obtain a clear picture of urban pollution.The aim of this thesis is to analyze and develop statistical models that exploit data acquired by low-cost mobile sensors. It contributes to the objective of providing more accurate spatial and temporal estimates and forecasts of urban air pollution, by combining mathematical models and technological advances
Faye, Papa Abdoulaye. "Planification et analyse de données spatio-temporelles." Thesis, Clermont-Ferrand 2, 2015. http://www.theses.fr/2015CLF22638/document.
Повний текст джерелаSpatio-temporal modeling allows to make the prediction of a regionalized variable at unobserved points of a given field, based on the observations of this variable at some points of field at different times. In this thesis, we proposed a approach which combine numerical and statistical models. Indeed by using the Bayesian methods we combined the different sources of information : spatial information provided by the observations, temporal information provided by the black-box and the prior information on the phenomenon of interest. This approach allowed us to have a good prediction of the variable of interest and a good quantification of incertitude on this prediction. We also proposed a new method to construct experimental design by establishing a optimality criterion based on the uncertainty and the expected value of the phenomenon
Hmamouche, Youssef. "Prédiction des séries temporelles larges." Electronic Thesis or Diss., Aix-Marseille, 2018. http://www.theses.fr/2018AIXM0480.
Повний текст джерелаNowadays, storage and data processing systems are supposed to store and process large time series. As the number of variables observed increases very rapidly, their prediction becomes more and more complicated, and the use of all the variables poses problems for classical prediction models.Univariate prediction models are among the first models of prediction. To improve these models, the use of multiple variables has become common. Thus, multivariate models and become more and more used because they consider more information.With the increase of data related to each other, the application of multivariate models is also questionable. Because the use of all existing information does not necessarily lead to the best predictions. Therefore, the challenge in this situation is to find the most relevant factors among all available data relative to a target variable.In this thesis, we study this problem by presenting a detailed analysis of the proposed approaches in the literature. We address the problem of prediction and size reduction of massive data. We also discuss these approaches in the context of Big Data.The proposed approaches show promising and very competitive results compared to well-known algorithms, and lead to an improvement in the accuracy of the predictions on the data used.Then, we present our contributions, and propose a complete methodology for the prediction of wide time series. We also extend this methodology to big data via distributed computing and parallelism with an implementation of the prediction process proposed in the Hadoop / Spark environment
Hmamouche, Youssef. "Prédiction des séries temporelles larges." Thesis, Aix-Marseille, 2018. http://www.theses.fr/2018AIXM0480.
Повний текст джерелаNowadays, storage and data processing systems are supposed to store and process large time series. As the number of variables observed increases very rapidly, their prediction becomes more and more complicated, and the use of all the variables poses problems for classical prediction models.Univariate prediction models are among the first models of prediction. To improve these models, the use of multiple variables has become common. Thus, multivariate models and become more and more used because they consider more information.With the increase of data related to each other, the application of multivariate models is also questionable. Because the use of all existing information does not necessarily lead to the best predictions. Therefore, the challenge in this situation is to find the most relevant factors among all available data relative to a target variable.In this thesis, we study this problem by presenting a detailed analysis of the proposed approaches in the literature. We address the problem of prediction and size reduction of massive data. We also discuss these approaches in the context of Big Data.The proposed approaches show promising and very competitive results compared to well-known algorithms, and lead to an improvement in the accuracy of the predictions on the data used.Then, we present our contributions, and propose a complete methodology for the prediction of wide time series. We also extend this methodology to big data via distributed computing and parallelism with an implementation of the prediction process proposed in the Hadoop / Spark environment
Ravel, Sabrina. "Implication des neurones à activité tonique du stiatum dans les processus cognitifs et motivationnels : Etude électrophysiologique de l'influence de la prédiction temporelle et de la signification affective des stimuli chez le singe." Aix-Marseille 2, 2001. http://www.theses.fr/2001AIX22004.
Повний текст джерелаCherdo, Yann. "Détection d'anomalie non supervisée sur les séries temporelle à faible coût énergétique utilisant les SNNs." Electronic Thesis or Diss., Université Côte d'Azur, 2024. http://www.theses.fr/2024COAZ4018.
Повний текст джерелаIn the context of the predictive maintenance of the car manufacturer Renault, this thesis aims at providing low-power solutions for unsupervised anomaly detection on time-series. With the recent evolution of cars, more and more data are produced and need to be processed by machine learning algorithms. This processing can be performed in the cloud or directly at the edge inside the car. In such a case, network bandwidth, cloud services costs, data privacy management and data loss can be saved. Embedding a machine learning model inside a car is challenging as it requires frugal models due to memory and processing constraints. To this aim, we study the usage of spiking neural networks (SNNs) for anomaly detection, prediction and classification on time-series. SNNs models' performance and energy costs are evaluated in an edge scenario using generic hardware models that consider all calculation and memory costs. To leverage as much as possible the sparsity of SNNs, we propose a model with trainable sparse connections that consumes half the energy compared to its non-sparse version. This model is evaluated on anomaly detection public benchmarks, a real use-case of anomaly detection from Renault Alpine cars, weather forecasts and the google speech command dataset. We also compare its performance with other existing SNN and non-spiking models. We conclude that, for some use-cases, spiking models can provide state-of-the-art performance while consuming 2 to 8 times less energy. Yet, further studies should be undertaken to evaluate these models once embedded in a car. Inspired by neuroscience, we argue that other bio-inspired properties such as attention, sparsity, hierarchy or neural assemblies dynamics could be exploited to even get better energy efficiency and performance with spiking models. Finally, we end this thesis with an essay dealing with cognitive neuroscience, philosophy and artificial intelligence. Diving into conceptual difficulties linked to consciousness and considering the deterministic mechanisms of memory, we argue that consciousness and the self could be constitutively independent from memory. The aim of this essay is to question the nature of humans by contrast with the ones of machines and AI
Tengku, Mohd Azahar Tuan Dir. "Génération de prédiction par la combinaison de fusion de données et de modélisation spatio-temporelle : application à la localisation de la répartition de la maladie basal stem rot dans les plantations de palmiers à huile." Thesis, La Rochelle, 2012. http://www.theses.fr/2012LAROS386.
Повний текст джерелаThis thesis represents a new approach for predicting plant disease in a plantation through combination of data fusion and spatio-temporal modelling. Plant disease is a major problem in the world of agriculture. Example in Malaysia, basalstem rot disease (BSR) caused by Ganoderma Boinense is the most serious disease for oil palm plantation in Malaysia. The fungus infects oil palm trees, initially causing yield loss and finally killing the trees. Various factors were previously reported to influence incidence of BSR, such as previous crops, techniques for replanting, types of soils and the age of trees. At present effective and sustainable management strategies to control BSR are hampered mainly by a lack of understanding of mechanisms of disease establishment, development and spread. The present research is an attempt to apply data fusion technique and temporal modelling in Geographical Information System (GIS) to investigate the behaviour of plant disease in a specific area (small skill area). This research will focus on how GIS can help to assess the distribution plant disease in a small scale plantation. With concurrent advances in global positioning systems (GPS) and the use of geographical Information Systems(GIS) techniques have provided powerful analysis tools for precision agriculture. Data for analysis were obtained from oil palm planting density experiments at MPOB research stations at Teluk Intan, Perak, Malaysia. In the case of BSR disease, the results of the predictive modelling show a significance correlation between predicted BSR diseases with visually observed BSR data. It found that the proposed predictive modelling has well predicted the presence of BSR disease. Although at the beginning stage of BSR diseases infection, the model has not fitted exactly the distribution of the disease, we believe that with the proper selection of the source of data, the performance of the model will be improved.Overall, the model has well predicted the presence of diseases with accuracy up to 98.9%
Es-Seddiqi, Mouna. "Le rôle de la voie amygdalo-nigro-striée dans les processus attentionnels dans les apprentissages instrumentaux, classiques et temporels." Thesis, Paris 6, 2017. http://www.theses.fr/2017PA066072/document.
Повний текст джерелаAssociative learning is a highly complex mechanism, involving several processes at the same time. The attentional process is one of the first to be mobilized during an association; it would also be involved to extract the temporal parameters associated with an unconditional biologically meaningful stimulus even before any effective association (Balsam, Drew, and Yang 2002). Some studies have shown the involvement of certain neurobiological structures, which may underlie attentional processes. For the Holland PC team, for example, orientation responses to a conditioned stimulus (top-down attention) (Lee et al., 2005) involve the central nucleus and nigro-striatal dopaminergic projections, whereas presentation of a new stimulus during an association (bottom-up attention) would rather imply the substancia inominata which would be modulated by the central nucleus of amygdala (CeA) and the parietal cortex (Holland and Gallagher 2006). At the same time, in temporal discrimination in which associative learning requires, besides discrete sensory stimuli, performances related to the judgment of durations, the mechanism of the attentional process mobilizes other conceptual models that gravitate mainly around the internal clock model and, in particular, the striatal beat frequency model which propose also neurobiological explanations (Matell & Meck, 2004). In this work, we aimed at understanding the role of the Amygdalo-nigro-striatal (ANS) circuit in the development of the attentional process in associative learning oriented towards discrete and temporal sensory stimuli in the rat. We also aimed at examining the role of this circuit in the evolution of the attentional process after over-training permitting the development of habits. In order to achieve this objective, we compared performance of rats with cross-lesion by altering the CeA in one hemisphere and the nigro-striatal circuit in the other hemisphere (Amygdalo-nigro-striatal disconnection; Contra group) to rats with lesions in the same hemisphere (CeA and nigro-striatal circuit: group Ipsi). A third group was submitted to bilateral lesions of the CeA (Amy group). A control group had pseudo lesions (groupe Sham).Through our three experimental groups (Contra, Ipsi and Amy) and the control group (Sham), we have shown the involvement of the CeA in the modulation of the attentional process when a novelty was introduced in the experimental situation (surprise) both in the presence of an appetitive discrete sensory stimulus and of a temporal stimulus in an aversive context. We have also shown that the ANS circuit is involved in habit formation and that there is probably a differential effect between the posterior and anterior part of the CeA. Our work also highlighted the implication of the nigro-striatal circuit in temporal discrimination and of the ANS circuit in the attentional treatment in temporal perception tasks, this effect being different depending on whether the discrimination concerns short or long durations
Ben, Soussia Amal. "Analyse prédictive des données d’apprentissage, en situation d’enseignement à distance." Electronic Thesis or Diss., Université de Lorraine, 2022. http://www.theses.fr/2022LORR0216.
Повний текст джерелаOver the past few decades, the adoption of e-learning has evolved rapidly and its use has been pushedeven further with the COVID-19 pandemic. The objective of this learning mode is to guarantee thecontinuity of the learning process. However, the online learning is facing several challenges, and themost widespread is the high failure rates among learners. This issue is due to many reasons such asthe heterogeneity of the learners and the diversity of their learning behaviors, their total autonomy, thelack and/or the inefficiency of the pedagogical provided follow-up. . .. Therefore, teachers need a systembased on analytical and intelligent methods allowing them an accurate and early prediction of at-risk offailure learners. This solution is commonly adopted in the state of the art. However, the work carried outdoes not respond to some particularities of the learning process (the continuity and evolution of learning,the diversity of learners and their total autonomy) and to some teachers expectations such as the alertgeneration.This thesis belongs to the field of learning analytics and uses the numeric traces of online learnersto design a predictive system (Early Warning Systems (EWS)) dedicated to teachers in online establish-ments. The objective of this EWS is to identify learners at risk as soon as possible in order to alertteachers about them. In order to achieve this objective, we have dealt with several sub-problems whichhave allowed us to elaborate four scientific contributions.We start by proposing an in-depth methodology based on the Machine Learning (ML) steps and thatallows the identification of four learning indicators among : performance, engagement, reactivity andregularity. This methodology also highlights the importance of temporal data for improving predictionperformance. In addition, this methodology allowed to define the model with the best ability to identifyat-risk learners.The 2nd contribution consists in proposing a temporal evaluation of the EWS using temporal metricswhich measure the precocity of the predictions and the stability of the system. From these two metrics,we study the trade-offs that exist between ML precision metrics and temporal metrics.Online learners are characterized by the diversity of their learning behaviors. Thus, an EWS shouldrespond to this diversity by ensuring an equitable functioning with the different learners profiles. Wepropose an evaluation methodology based on the identification of learner profiles and that uses a widespectrum of temporal and precision metrics.By using an EWS, teachers expect an alert generation. For this reason, we design an algorithm which,based on the results of the prediction, the temporal metrics and the notion of alert rules, proposes anautomatic method for alert generation. This algorithm targets mainly at-risk learners.The context of this thesis is the French National Center for Distance Education (CNED). In parti-cular, we use the numeric traces of k-12 learners enrolled during the 2017-2018 and 2018-2019 schoolyears
Senouci, Sid-Ahmed Benali. "Optimisation et prédiction temporelles sur les réseaux programmables CPLD." Ecully, Ecole centrale de Lyon, 1998. http://www.theses.fr/1998ECDL0051.
Повний текст джерелаChleq, Nicolas. "Contribution à l'étude du raisonnement temporel : résolution avec contraintes et application à l'abduction en raisonnement temporel." Phd thesis, Ecole Nationale des Ponts et Chaussées, 1995. http://tel.archives-ouvertes.fr/tel-00529412.
Повний текст джерелаCissoko, Mamadou Ben Hamidou. "Adaptive time-aware LSTM for predicting and interpreting ICU patient trajectories from irregular data." Electronic Thesis or Diss., Strasbourg, 2024. http://www.theses.fr/2024STRAD012.
Повний текст джерелаIn personalized predictive medicine, accurately modeling a patient's illness and care processes is crucial due to the inherent long-term temporal dependencies. However, Electronic Health Records (EHRs) often consist of episodic and irregularly timed data, stemming from sporadic hospital admissions, which create unique patterns for each hospital stay. Consequently, constructing a personalized predictive model necessitates careful consideration of these factors to accurately capture the patient's health journey and assist in clinical decision-making. LSTM networks are effective for handling sequential data like EHRs, but they face two significant limitations: the inability to interpret prediction results and to take into account irregular time intervals between consecutive events. To address limitations, we introduce novel deep dynamic memory neural networks called Multi-Way Adaptive and Adaptive Multi-Way Interpretable Time-Aware LSTM (MWTA-LSTM and AMITA) designed for irregularly collected sequential data. The primary objective of both models is to leverage medical records to memorize illness trajectories and care processes, estimate current illness states, and predict future risks, thereby providing a high level of precision and predictive power
Zuo, Jingwei. "Apprentissage de représentations et prédiction pour des séries-temporelles inter-dépendantes." Electronic Thesis or Diss., université Paris-Saclay, 2022. http://www.theses.fr/2022UPASG038.
Повний текст джерелаTime series is a common data type that has been applied to enormous real-life applications, such as financial analysis, medical diagnosis, environmental monitoring, astronomical discovery, etc. Due to its complex structure, time series raises several challenges in their data processing and mining. The representation of time series plays a key role in data mining tasks and machine learning algorithms for time series. Yet, a few methods consider the interrelation that may exist between different time series when building the representation. Moreover, the time series mining requires considering not only the time series' characteristics in terms of data complexity but also the concrete application scenarios where the data mining task is performed to build task-specific representations.In this thesis, we will study different time series representation approaches that can be used in various time series mining tasks, while capturing the relationships among them. We focus specifically on modeling the interrelations between different time series when building the representations, which can be the temporal relationship within each data source or the inter-variable relationship between various data sources. Accordingly, we study the time series collected from various application contexts under different forms. First, considering the temporal relationship between the observations, we learn the time series in a dynamic streaming context, i.e., time series stream, for which the time series data is continuously generated from the data source. Second, for the inter-variable relationship, we study the multivariate time series (MTS) with data collected from multiple data sources. Finally, we study the MTS in the Smart City context, when each data source is given a spatial position. The MTS then becomes a geo-located time series (GTS), for which the inter-variable relationship requires more modeling efforts with the external spatial information. Therefore, for each type of time series data collected from distinct contexts, the interrelations between the time series observations are emphasized differently, on the temporal or (and) variable axis.Apart from the data complexity from the interrelations, we study various machine learning tasks on time series in order to validate the learned representations. The high-level learning tasks studied in this thesis consist of time series classification, semi-supervised time series learning, and time series forecasting. We show how the learned representations connect with different time series learning tasks under distinct application contexts. More importantly, we conduct the interdisciplinary study on time series by leveraging real-life challenges in machine learning tasks, which allows for improving the learning model's performance and applying more complex time series scenarios.Concretely, for these time series learning tasks, our main research contributions are the following: (i) we propose a dynamic time series representation learning model in the streaming context, which considers both the characteristics of time series and the challenges in data streams. We claim and demonstrate that the Shapelet, a shape-based time series feature, is the best representation in such a dynamic context; (ii) we propose a semi-supervised model for representation learning in multivariate time series (MTS). The inter-variable relationship over multiple data sources is modeled in a real-life context, where the data annotations are limited; (iii) we design a geo-located time series (GTS) representation learning model for Smart City applications. We study specifically the traffic forecasting task, with a focus on the missing-value treatment within the forecasting algorithm
Ziat, Ali Yazid. "Apprentissage de représentation pour la prédiction et la classification de séries temporelles." Thesis, Paris 6, 2017. http://www.theses.fr/2017PA066324/document.
Повний текст джерелаThis thesis deals with the development of time series analysis methods. Our contributions focus on two tasks: time series forecasting and classification. Our first contribution presents a method of prediction and completion of multivariate and relational time series. The aim is to be able to simultaneously predict the evolution of a group of time series connected to each other according to a graph, as well as to complete the missing values in these series (which may correspond for example to a failure of a sensor during a given time interval). We propose to use representation learning techniques to forecast the evolution of the series while completing the missing values and taking into account the relationships that may exist between them. Extensions of this model are proposed and described: first in the context of the prediction of heterogeneous time series and then in the case of the prediction of time series with an expressed uncertainty. A prediction model of spatio-temporal series is then proposed, in which the relations between the different series can be expressed more generally, and where these can be learned.Finally, we are interested in the classification of time series. A joint model of metric learning and time-series classification is proposed and an experimental comparison is conducted
Ziat, Ali Yazid. "Apprentissage de représentation pour la prédiction et la classification de séries temporelles." Electronic Thesis or Diss., Paris 6, 2017. http://www.theses.fr/2017PA066324.
Повний текст джерелаThis thesis deals with the development of time series analysis methods. Our contributions focus on two tasks: time series forecasting and classification. Our first contribution presents a method of prediction and completion of multivariate and relational time series. The aim is to be able to simultaneously predict the evolution of a group of time series connected to each other according to a graph, as well as to complete the missing values in these series (which may correspond for example to a failure of a sensor during a given time interval). We propose to use representation learning techniques to forecast the evolution of the series while completing the missing values and taking into account the relationships that may exist between them. Extensions of this model are proposed and described: first in the context of the prediction of heterogeneous time series and then in the case of the prediction of time series with an expressed uncertainty. A prediction model of spatio-temporal series is then proposed, in which the relations between the different series can be expressed more generally, and where these can be learned.Finally, we are interested in the classification of time series. A joint model of metric learning and time-series classification is proposed and an experimental comparison is conducted
Arnoux, Thibaud. "Prédiction d'interactions dans les flots de liens. Combiner les caractéristiques structurelles et temporelles." Electronic Thesis or Diss., Sorbonne université, 2018. http://www.theses.fr/2018SORUS229.
Повний текст джерелаThe link stream formalism represent an approach allowing to capture the system dynamic while providing a framework to understand the system's behavior. A link stream is a sequence of triplet (t,u,v) indicating that an interaction occurred between u and v at time t. The importance of the system's dynamic during the prediction places it at the crossroads of link prediction in graphs and time series prediction. We will explore several formalizations of the problem of prediction in link streams. In the following we will study the activity prediction, that is to say predicting the number of interactions occurring in the future between each pair of nodes during a given period. We introduce the protocol, allowing to combine the data characteristics to predict the activity. We study the behavior of our protocol during several experiments on four datasets et evaluate the prediction quality. We will look at how the introduction of pair of nodes classes allows to preserve the link diversity in the prediction while improving the prediction. Our goal is to define a general prediction framework allowing in-depth studies of the relationship between temporal and structural characteristics in prediction tasks
Dzeutouo, Zapa Donard. "Développement d’un modèle prédictif de la productivité spatio-temporelle des plants de bleuets sauvages." Mémoire, Université de Sherbrooke, 2017. http://hdl.handle.net/11143/11331.
Повний текст джерелаSalamat, Nadeem. "Modélisation des relations spatiales entre objets en mouvement." Phd thesis, Université de La Rochelle, 2011. http://tel.archives-ouvertes.fr/tel-00718399.
Повний текст джерелаVirolle, Maxime. "Origine et prédiction spatio-temporelle des tapissages argileux dans les réservoirs silicoclastiques - Apports de la comparaison entre des réservoirs enfouis (Permien et Crétacé) et un analogue actuel (estuaire de la Gironde) Detrital clay grain coats in estuarine clastic deposits: origin and spatial distribution within a modern sedimentary system, the Gironde Estuary (south-west France) Influence of sedimentation and detrital clay grain coats on chloritized sandstone reservoir qualities: Insights from comparisons between ancient tidal heterolithic sandstones and a modern estuarine system Identification of a chloritization process in the Wealden facies sandstones (Early Cretaceous) of the Paris Basin, France." Thesis, Université Paris-Saclay (ComUE), 2019. http://www.theses.fr/2019SACLS190.
Повний текст джерелаThe reservoir quality is one of the "risk" factors for hydrocarbon exploration or for the future development of geothermal energy in siliciclastic hydrosystems. Reservoir properties are defined by porosity and permeability values. In deep buried siliciclastic reservoirs, chlorite coatings around quartz grains help to preserve these properties. The mechanisms behind these coatings are still poorly understood. The objectives of this study are: (1) to characterize (mineralogy, crystallography, textural and microstructural properties) and to determine the spatial and temporal distribution of clay and clay coatings in well constrained sedimentary environments and within a well-defined stratigraphic framework; (2) to better understand the factors controlling the formation of clay coatings in siliciclastic reservoirs; (3) to describe the intermediate processes of authigenic chlorite formation via different precursor minerals during burial; (4) to predict the distribution of good reservoir properties in relation to clay coatings. The modern analogue chosen for this study is the Gironde estuary, where the presence of detrital clay grain coats was detected in the intertidal zone of tidal and point bars, but also in pluri-meters long sedimentary cores. The formation mechanisms of these coatings have been investigated with the interaction between hydrodynamic and biological processes. Analogies with buried sandstone reservoirs (>3500m) showed that sand facies of external tidal bars deposited at the end of a transgressive cycle are the preferred targets for finding good reservoir properties in estuarine reservoirs. The evolution of detrital clay grain coats during burial was approached by studying buried reservoirs between 400 and 1000m deep. The detrital coatings are transformed into berthierine and mixed-layer chlorite-smectite at depths between 600 and 900m and temperatures between 30-40°C. These minerals are true precursors to ferrous chlorite coatings that appear at greater depth
Vroman, Philippe. "Prédiction des séries temporelles en milieu incertain : application à la prévision de ventes dans la distribution textile." Lille 1, 2000. http://www.theses.fr/2000LIL10207.
Повний текст джерелаSalaün, Achille. "Prédiction d'alarmes dans les réseaux via la recherche de motifs spatio-temporels et l'apprentissage automatique." Electronic Thesis or Diss., Institut polytechnique de Paris, 2021. http://www.theses.fr/2021IPPAS010.
Повний текст джерелаNowadays, telecommunication networks occupy a central position in our world. Indeed, they allow to share worldwide a huge amount of information. Networks are however complex systems, both in size and technological diversity. Therefore, it makes their management and reparation more difficult. In order to limit the negative impact of such failures, some tools have to be developed to detect a failure whenever it occurs, analyse its root causes to solve it efficiently, or even predict this failure as prevention is better than cure. In this thesis, we mainly focus on these two last problems. To do so, we use files, called alarm logs, storing all the alarms that have been emitted by the system. However, these files are generally noisy and verbose: an operator managing a network needs tools able to extract and handle in an interpretable manner the causal relationships inside a log. In this thesis, we followed two directions. First, we have inspired from pattern matching techniques: similarly to the Ukkonen’s algorithm, we build online a structure, called DIG-DAG, that stores all the potential causal relationships between the events of a log. Moreover, we introduce a query system to exploit our DIG-DAG structure. Finally, we show how our solution can be used for root cause analysis. The second approach is a generative approach for the prediction of time series. In particular, we compare two well-known models for this task: recurrent neural nets on the one hand, hidden Markov models on the other hand. Here, we compare analytically the expressivity of these models by encompassing them into a probabilistic model, called GUM
Nguyen, Thi Thu Tam. "Learning techniques for the load forecasting of parcel pick-up points." Electronic Thesis or Diss., université Paris-Saclay, 2023. http://www.theses.fr/2023UPASG034.
Повний текст джерелаPick-Up Points (PUP) represent an alternative delivery option for purchases from online retailers (Business-to-Customer, B2C) or online Customer-to-Customer (C2C) marketplaces. Parcels are delivered at a reduced cost to a PUP and wait until being picked up by customers or returned to the original warehouse if their sojourn time is over. When the chosen PUP is overloaded, the parcel may be refused and delivered to the next available PUP on the carrier tour. PUP load forecasting is an efficient method for the PUP management company (PMC) to better balance the load of each PUP and reduce the number of rerouted parcels. This thesis aims to describe the parcel flows in a PUP and to proposed models used to forecast the evolution of the load. For the PUP load associated with the B2C business, the parcel life-cycle has been taken into account in the forecasting process via models of the flow of parcel orders, the delivery delays, and the pick-up process. Model-driven and data-driven approaches are compared in terms of load-prediction accuracy. For the PUP load associated with the C2C business, the daily number of parcels dropped off with a given PUP as target is described by a Markov-Switching AutoRegressive model to account for the non-stationarity of the second-hand shopping activity. The life-cycle of each parcel is modeled by a Markov jump process. Model parameters are evaluated from previous parcel drop-off, delivery, and pick-up records. The probability mass function of the future load of a PUP is then evaluated using all information available on parcels with this PUP as target. In both cases, the proposed model-driven approaches give, for most of the cases, better forecasting performance, compared with the data-driven models, involving LSTM, Random forest, Holt-Winters, and SARIMA models, up to four days ahead in the B2C case and up to six days ahead in the C2C case. The first approach applied to the B2C parcel load yields an MAE of 3 parcels for the one-day ahead prediction and 8 parcels for the four-day ahead prediction. The second approach applied to the C2C parcel load yields an MAE of 5 parcels for the one-day ahead prediction and 8 parcels for the seven-day ahead prediction. These prediction horizons are consistent with the delivery delay associated with these parcels (1-3 days in the case of a B2C parcel and 4-5 days in the case of a C2C parcel). Future research directions aim at optimizing the prediction accuracy, especially in predicting future orders and studying a load-balancing approach to better share the load between PUPs
Han, Biao. "Predictive coding : its spike-time based neuronal implementation and its relationship with perception and oscillations." Thesis, Toulouse 3, 2016. http://www.theses.fr/2016TOU30029/document.
Повний текст джерелаIn this thesis, we investigated predictive coding and its relationship with perception and oscillations. We first reviewed my current understanding about facts of neuron and neocortex and state-of-the-arts of predictive coding in the introduction. In the main chapters, firstly, we proposed the idea that correlated spike times create selective inhibition in a nonselective excitatory feedback network in a theoretical study. Then, we showed the perceptual effect of predictive coding: shape perception enhances perceived contrast. At last, we showed that predictive coding can use oscillations with different frequencies for feedforward and feedback. This thesis provided an innovative and viable neuronal mechanism for predictive coding and empirical evidence for excitatory predictive feedback and the close relationship between the predictive coding and oscillations
Lepère, Stéphane. "Contribution à la prédiction en ligne des séries temporelles : un cas d'étude à la modélisation de systèmes dynamiques." Lille 1, 2001. https://pepite-depot.univ-lille.fr/RESTREINT/Th_Num/2001/50376-2001-219.pdf.
Повний текст джерелаKhodor, Nadine. "Analyse de la dynamique des séries temporelles multi-variées pour la prédiction d’une syncope lors d’un test d’inclinaison." Thesis, Rennes 1, 2014. http://www.theses.fr/2014REN1S123/document.
Повний текст джерелаSyncope is a sudden loss of consciousness. Although it is not usually fatal, it has an economic impact on the health care system and the personal lives of people suffering. The purpose of this study is to reduce the duration of the clinical test (approximately 1 hour) and to avoid patients to develop syncope by early predicting the occurrence of syncope. The entire work fits into a data mining approach involving the feature extraction, feature selection and classification. 3 complementary approaches are proposed, the first one exploits nonlinear analysis methods of time series extracted from signals acquired during the test, the second one focuses on time- frequency (TF) relation between signals and suggests new indexes and the third one, the most original, takes into account their temporal dynamics
Lefieux, Vincent. "Modèles semi-paramétriques appliqués à la prévision des séries temporelles. Cas de la consommation d'électricité." Phd thesis, Université Rennes 2, 2007. http://tel.archives-ouvertes.fr/tel-00179866.
Повний текст джерелаMarquez, Alfonzo Bicky. "Reservoir computing photonique et méthodes non-linéaires de représentation de signaux complexes : Application à la prédiction de séries temporelles." Thesis, Bourgogne Franche-Comté, 2018. http://www.theses.fr/2018UBFCD042/document.
Повний текст джерелаArtificial neural networks are systems prominently used in computation and investigations of biological neural systems. They provide state-of-the-art performance in challenging problems like the prediction of chaotic signals. Yet, the understanding of how neural networks actually solve problems like prediction remains vague; the black-box analogy is often employed. Merging nonlinear dynamical systems theory with machine learning, we develop a new concept which describes neural networks and prediction within the same framework. Taking profit of the obtained insight, we a-priori design a hybrid computer, which extends a neural network by an external memory. Furthermore, we identify mechanisms based on spatio-temporal synchronization with which random recurrent neural networks operated beyond their fixed point could reduce the negative impact of regular spontaneous dynamics on their computational performance. Finally, we build a recurrent delay network in an electro-optical setup inspired by the Ikeda system, which at first is investigated in a nonlinear dynamics framework. We then implement a neuromorphic processor dedicated to a prediction task
Voyant, Cyril. "Prédiction de séries temporelles de rayonnement solaire global et de production d'énergie photovoltaïque à partir de réseaux de neurones artificiels." Phd thesis, Université Pascal Paoli, 2011. http://tel.archives-ouvertes.fr/tel-00635298.
Повний текст джерелаWacongne, Catherine. "Traitements conscient et non-conscient des régularités temporelles : Modélisation et neuroimagerie." Thesis, Paris 6, 2014. http://www.theses.fr/2014PA066290/document.
Повний текст джерелаWhat is going to happen next? Natural stimuli tend to follow each other in a reproducible way. Multiple fields of neuroscience and psychology bring evidence that human’s brain and behavior are sensitive to the temporal structure of stimuli and are able to exploit them in multiple ways: to make appropriate decisions, encode efficiently information, react faster to predictable stimuli or orient attention towards surprising ones… Multiple brain areas show sensitivity to the temporal structure of events. However, all areas do not seem to be sensitive to the same kind of temporal regularities. Conscious access to the stimuli seems to play a key role in some of these dissociations and better understanding this role could improve the current diagnostic tools for non-communicative patients. This thesis explores the hierarchical organization of the processing of temporal regularities and the computational properties of conscious and unconscious levels of processing by combining a modeling approach with neuroimaging experiments using magnetoencephalography and electroencephalography (MEEG). First, a plausible neuronal model based on predictive coding principles reproduces the main properties of the preattentive processing of pure tones in the auditory cortex indexed by the evoked potential mismatch negativity (MMN). Second, a MEEG experiment provides evidence for a hierarchical organization of multiple predictive processes in the auditory cortex. Finally, a second model explores the new computational properties and constraints associated to the access of stimuli to a conscious space with a working memory able to maintain information for an arbitrary time but with limited capacity
Germain, Simon. "Conception d'une mesure automatisée de détection des changements alimentaires chez le porc." Mémoire, Université de Sherbrooke, 2015. http://hdl.handle.net/11143/7925.
Повний текст джерелаVuillemin, Benoit. "Recherche de règles de prédiction dans un contexte d'Intelligence Ambiante." Thesis, Lyon, 2020. http://www.theses.fr/2020LYSE1120.
Повний текст джерелаThis thesis deals with the subject of Ambient Intelligence, the fusion between Artificial Intelligence and the Internet of Things. The goal of this work is to extract prediction rules from the data provided by connected objects in an environment, in order to propose automation to users. Our main concern relies on privacy, user interactions, and the explainability of the system’s operation. In this context, several contributions were made. The first is an ambient intelligence architecture that operates locally, and processes data from a single connected environment. The second is a discretization process without a priori on the input data, allowing to take into account different kinds of data from various objects. The third is a new algorithm for searching rules over a time series, which avoids the limitations of stateoftheart algorithms. The approach was validated by tests on two real databases. Finally, prospects for future developments in the system are presented
Çinar, Yagmur Gizem. "Prédiction de séquences basée sur des réseaux de neurones récurrents dans le contexte des séries temporelles et des sessions de recherche d'information." Thesis, Université Grenoble Alpes (ComUE), 2019. http://www.theses.fr/2019GREAM079.
Повний текст джерелаThis thesis investigates challenges of sequence prediction in different scenarios such as sequence prediction using recurrent neural networks (RNNs) in the context of time series and information retrieval (IR) search sessions. Predicting the unknown values that follow some previously observed values is basically called sequence prediction.It is widely applicable to many domains where a sequential behavior is observed in the data. In this study, we focus on two different types of sequence prediction tasks: time series forecasting and next query prediction in an information retrieval search session.Time series often display pseudo-periods, i.e. time intervals with strong correlation between values of time series. Seasonal changes in weather time series or electricity usage at day and night time are some examples of pseudo-periods. In a forecasting scenario, pseudo-periods correspond to the difference between the positions of the output being predicted and specific inputs.In order to capture periods in RNNs, one needs a memory of the input sequence. Sequence-to-sequence RNNs (with attention mechanism) reuse specific (representations of) input values to predict output values. Sequence-to-sequence RNNs with an attention mechanism seem to be adequate for capturing periods. In this manner, we first explore the capability of an attention mechanism in that context. However, according to our initial analysis, a standard attention mechanism did not perform well to capture the periods. Therefore, we propose a period-aware content-based attention RNN model. This model is an extension of state-of-the-art sequence-to-sequence RNNs with attention mechanism and it is aimed to capture the periods in time series with or without missing values.Our experimental results with period-aware content-based attention RNNs show significant improvement on univariate and multivariate time series forecasting performance on several publicly available data sets.Another challenge in sequence prediction is the next query prediction. The next query prediction helps users to disambiguate their search query, to explore different aspects of the information they need or to form a precise and succint query that leads to higher retrieval performance. A search session is dynamic, and the information need of a user might change over a search session as a result of the search interactions. Furthermore, interactions of a user with a search engine influence the user's query reformulations. Considering this influence on the query formulations, we first analyze where the next query words come from? Using the analysis of the sources of query words, we propose two next query prediction approaches: a set view and a sequence view.The set view adapts a bag-of-words approach using a novel feature set defined based on the sources of next query words analysis. Here, the next query is predicted using learning to rank. The sequence view extends a hierarchical RNN model by considering the sources of next query words in the prediction. The sources of next query words are incorporated by using an attention mechanism on the interaction words. We have observed using sequence approach, a natural formulation of the problem, and exploiting all sources of evidence lead to better next query prediction
Diss, Stéphanie. "Apport de l'imagerie radar pour la connaissance spatio-temporelle des champs de pluie : utilisation pour une modélisation prédictive des crues." Paris 6, 2009. http://www.theses.fr/2009PA066038.
Повний текст джерелаRynkiewicz, Joseph. "Modèles hybrides intégrant des réseaux de neurones artificiels à des modèles de chaînes de Markov cachées : application à la prédiction de séries temporelles." Paris 1, 2000. http://www.theses.fr/2000PA010077.
Повний текст джерелаLefieux, Vincent. "Modèles semi-paramétriques appliqués à la prévision des séries temporelles : cas de la consommation d’électricité." Phd thesis, Rennes 2, 2007. https://theses.hal.science/tel-00179866/fr/.
Повний текст джерелаRéseau de Transport d’Electricité (RTE), in charge of operating the French electric transportation grid, needs an accurate forecast of the power consumption in order to operate it correctly. The forecasts used everyday result from a model combining a nonlinear parametric regression and a SARIMA model. In order to obtain an adaptive forecasting model, nonparametric forecasting methods have already been tested without real success. In particular, it is known that a nonparametric predictor behaves badly with a great number of explanatory variables, what is commonly called the curse of dimensionality. Recently, semiparametric methods which improve the pure nonparametric approach have been proposed to estimate a regression function. Based on the concept of ”dimension reduction”, one those methods (called MAVE : Moving Average -conditional- Variance Estimate) can apply to time series. We study empirically its effectiveness to predict the future values of an autoregressive time series. We then adapt this method, from a practical point of view, to forecast power consumption. We propose a partially linear semiparametric model, based on the MAVE method, which allows to take into account simultaneously the autoregressive aspect of the problem and the exogenous variables. The proposed estimation procedure is practicaly efficient
Nguyen, Viet Hoa. "Une méthode fondée sur les modèles pour gérer les propriétés temporelles des systèmes à composants logiciels." Thesis, Rennes 1, 2013. http://www.theses.fr/2013REN1S090/document.
Повний текст джерелаThis thesis proposes an approach to integrate the use of time-related stochastic properties in a continuous design process based on models at runtime. Time-related specification of services are an important aspect of component-based architectures, for instance in distributed, volatile networks of computer nodes. The models at runtime approach eases the management of such architectures by maintaining abstract models of architectures synchronized with the physical, distributed execution platform. For self-adapting systems, prediction of delays and throughput of a component assembly is of utmost importance to take adaptation decision and accept evolutions that conform to the specifications. To this aim we define a metamodel extension based on stochastic Petri nets as an internal time model for prediction. We design a library of patterns to ease the specification and prediction of common time properties of models at runtime and make the synchronization of behaviors and structural changes easier. Furthermore, we apply the approach of Aspect-Oriented Modeling to weave the internal time models into timed behavior models of the component and the system. Our prediction engine is fast enough to perform prediction at runtime in a realistic setting and validate models at runtime
Cason, Nia. "L'effet du rythme musical sur la parole." Thesis, Aix-Marseille, 2013. http://www.theses.fr/2013AIXM4086/document.
Повний текст джерелаMusic and speech are both reliant on how events occur in time. Both require anticipation about when and what events will occur as well as a temporal and hierarchical organisation of salient and less salient events. These may rely on common, domain-general processes.With this in mind, three experiments using behavioural and electrophysiological (EEG) measures were conducted which aimed to investigate whether speech perception and production can benefit from rhythmic priming (inducing temporal expectations through music, and which can inform a listener about temporal structures in speech). We have found that phonological processing of spoken pseudowords is enhanced when speech conforms to listener expectations, as measured by behavioural (reaction time) and EEG data (Cason & Schön, 2012). Phonological processing of sentences can also be enhanced via rhythmic priming (behavioural measures) and this priming effect is augmented through training with the musical rhythms (Cason, Astésano & Schön, submitted).Overall, it seems that the regularity of musical rhythm (over speech rhythm) allows a listener to form precise temporal expectations and a metrical memory trace which can impact on phonological processing of words and sentences, and that rhythmic priming can also enhance articulation performance in hearing-impaired children, perhaps via an enhanced phonological perception
Mecharnia, Thamer. "Approches sémantiques pour la prédiction de présence d'amiante dans les bâtiments : une approche probabiliste et une approche à base de règles." Electronic Thesis or Diss., université Paris-Saclay, 2022. http://www.theses.fr/2022UPASG036.
Повний текст джерелаNowadays, Knowledge Graphs are used to represent all kinds of data and they constitute scalable and interoperable resources that can be used by decision support tools. The Scientific and Technical Center for Building (CSTB) was asked to develop a tool to help identify materials containing asbestos in buildings. In this context, we have created and populated the ASBESTOS ontology which allows the representation of building data and the results of diagnostics carried out in order to detect the presence of asbestos in the used products. We then relied on this knowledge graph to develop two approaches which make it possible to predict the presence of asbestos in products in the absence of the reference of the marketed product actually used.The first approach, called the hybrid approach, is based on external resources describing the periods when the marketed products are asbestos-containing to calculate the probability of the existence of asbestos in a building component. This approach addresses conflicts between external resources, and incompleteness of listed data by applying a pessimistic fusion approach that adjusts the calculated probabilities using a subset of diagnostics.The second approach, called CRA-Miner, is inspired by inductive logic programming (ILP) methods to discover rules from the knowledge graph describing buildings and asbestos diagnoses. Since the reference of specific products used during construction is never specified, CRA-Miner considers temporal data, ASBESTOS ontology semantics, product types and contextual information such as part-of relations to discover a set of rules that can be used to predict the presence of asbestos in construction elements.The evaluation of the two approaches carried out on the ASBESTOS ontology populated with the data provided by the CSTB show that the results obtained, in particular when the two approaches are combined, are quite promising
Ramanantenasoa, Maharavo. "Prise en compte de la variabilité spatio-temporelle des émissions d'ammoniac liées à la fertilisation azotée en France et développement de métamodèles prédictifs." Thesis, Université Paris-Saclay (ComUE), 2018. http://www.theses.fr/2018SACLA027/document.
Повний текст джерелаIn a context of reducing the impacts of agricultural practices on human health andecosystems, it is necessary to better account for ammonia (NH3) volatilization in the inventories of NH3 emissions, the understanding of the nitrogen behavior after fertilization and the modeling of air quality.Given the considerable weight of nitrogen fertilizer (55%) in the total national NH3 emissions, nitrogen fertilization is an important lever for reducing NH3 emissions. Current national inventories are based on the use of default emission factors (EF) and suffer from a lack of fine spatial and temporal NH3 emissions descriptions making it difficult to develop effective emission reduction policies. Moreover, even if there are currently models that globally simulate the fate of nitrogen on the crop cycle, their do not always take into account the volatilization of NH3, and if it is, the volatilization modules are often very frustrating and have not been validated. There are models dedicated exclusively to the volatilization of NH3 in the field, but their requirements for data and input parameters and their calculation time limit their large-scale use (e.g, at national scale) in many geographical locations for several years as well as their integration into airquality prediction models or decision support or environmental assessment tools in terms of nitrogen fertilization.This thesis proposes new tools. The first tool, CADASTRE_NH3 makes it possible to describe and analyze the spatio-temporal variability of NH3 emissions from nitrogen fertilization. It combines the Volt'Air process-based model with geo-referenced databases on agro-soil-meteorological factors at the national level. This approach has demonstrated its ability to capture the spatio-temporal variability ofnitrogen use and resulting NH3 emissions, and to take into account the effect of soil and climate factor interactions on emissions. The comparison of CADASTRE_NH3 results with the official French inventories shows strong convergences regarding the quantities of nitrogen used and NH3 emissions in France for the year 2005-06, but also divergences especially for the case of organic waste products. Thesecond type of tool corresponds to meta-models derived from Volt'Air for predicting NH3 emission rates after nitrogen fertilizer applications. These meta-models have many practical advantages because of their simplicity and operability. They have promising potential uses to support decision-making in terms of fertilizer use conditions and also to support emission reduction policies through, for example, scenario testing.However, it would be interesting to compare the tools developed in this thesis with experimental data to evaluate their respective performance and validate all our approaches. It would also be interesting to develop dynamic meta-models of Volt'Air that can describe the dynamics of NH3 fluxes related to nitrogen fertilizer applications and to be integrated as simple modules of NH3 volatilization in agronomic and atmospheric models
Dermouche, Soumia. "Leveraging the dynamics of non-verbal behaviors : modeling social attitude and engagement in human-agent interaction." Electronic Thesis or Diss., Sorbonne université, 2019. https://accesdistant.sorbonne-universite.fr/login?url=https://theses-intra.sorbonne-universite.fr/2019SORUS271.pdf.
Повний текст джерелаSocial interaction implies exchange between two or more persons, where they adapt their behaviors to each others. With the growing interest in human-agent interactions, it is desirable to make these interactions natural and human like. In this context, we aimed at enhancing the quality of the interaction between users and Embodied Conversational Agents ECAs by (1) endowing the ECA with the capacity to express social attitudes, such as being friendly or dominant depending its role or relationship with its interaction partners; (2) adapting the agent's behavior according to the user's behavior, hence, the conversation partners influence each others through an interaction loop, thus, enhancing the interaction quality; (3) predicting the user's engagement level and adapting the agent's behavior accordingly. We take advantage of the recent advances in machine learning, more specifically, temporal sequence mining and neural networks to model these capacities in the ECA. The first model is used to learn relevant patterns (sequences) of non-verbal signals that best represent attitude variations, and then reproduce them on the agent. The latter is used to encompass the dynamics of non-verbal signals. Two use cases have been explored using the well-known LSTM model: agent's behavior adaptation based on both agent's and user's behavior history, and user's engagement prediction based on his/her own behavior history. The implemented models and algorithms have been validated through a number of perceptive studies as well as through rigorous quantitative analysis of the obtained results. In addition, the realized models have been integrated into a virtual-agent platform
Wang, Chao. "Modélisation et prédiction des assemblages de phytoplancton à l’aval de la rivière des Perles, en Chine." Toulouse 3, 2014. http://thesesups.ups-tlse.fr/2666/.
Повний текст джерелаFreshwater ecosystems throughout the world are experiencing increasing pressures from both climate changes and anthropogenic activities. Rivers, the typical lotic freshwater ecosystems, are regarded as important pathways for the flow of energy, matter, and organisms through the landscape. Phytoplankton constitutes the base level of the aquatic food web, and it has quick response to environmental factors that regulate biological activity and water quality. Studies on phytoplankton have been extensive in lentic fresh-waters such as lakes and reservoirs, but still less in lotic ecosystems. The Pearl River is the largest lowland river of South China, but relevant studies were interrupted during the last three decades. Consequently in the present study, we contribute to highlight the patterns of the phytoplankton assemblages of this large river, with the approach of several ecological modeling. Firstly, we summarize the scientific trends in phytoplankton studies between 1991 and 2013 based on bibliometric analysis. Although the annual publication output of phytoplankton demonstrated a rapid linear increasing tendency during the last two decades, its contribution to total scientific articles always kept below 10%. Under the background of fast scientific research development, dependent publications (in terms of multi-aquatic ecosystems and international collaborations) indicate linear increasing trend. The variations of keywords associated with research regions are mostly impacted by the geographic adjacent countries, which are generally the top contributors. Variation trends of all the keywords relating to research methods, research contents and environmental factors indicate that phytoplankton studies carried out in large scale and long term are in significant ascending trend, while traditional and local scale studies are in descending trend. Secondly, temporal patterns of phytoplankton assemblages were analyzed within the downstream region of the Pearl River (China), through time-series sampling during the whole of 2009. The excessive nutrient conditions resulted in a diatom dominant phytoplankton community. While green algae only contributed more in species diversity. Phytoplankton samples were classified into four clusters using a self-organizing map (SOM) based on species similarities. These clusters were clearly different, with respect to species richness, biomass and indicators. Moreover, the LDA predicting model indicated that these clusters could easily be differentiated by physical factors such as water temperature, discharge and precipitation. As for nutrients, only phosphate could have an occasional impact on phytoplankton assemblages. The global score for predicting the assemblages was 64. 2%. Thirdly, spatial patterns of phytoplankton were analyzed within the Pearl River delta system (China), through seasonal sampling during 2012. The excessive nutrient conditions and well water exchanges resulted in a phytoplankton community that Bacillariophyceae and Chlorophyceae dominated in diversity and Bacillariophyceae dominated in biomass. Phytoplankton samples were revealed by the ordination method using a NMDS and five groups were determined by using hclust. These groups were clearly different, with respect to species richness, biomass and indicators, but differences between the patterning groups were only significant in spatial dimension. The LDA predicting model indicated that the spatial patterns of phytoplankton community assemblages could easily be differentiated by variables (TP, Si, DO and transparency) associated with water quality. The global score for predicting the assemblages was 75%. Lastly, the morphological variability of the predominant diatom species, Aulacoseira granulata (Ehrenberg) Simonsen, was observed within the downstream region of the Pearl River (China). High coherence between morphological parameters, especially cell size, was confirmed. Moreover, phase angles in wavelet figures also illustrated that cell diameter was the most sensitive parameter to environmental variations and through this way cell and filament size variations could be related. Water temperature impacted algal occurrence rates and size values during the spring-winter period. Algal life cycle could be affected by discharge, as well as filament length by allowing for selection of chains with optimum buoyancy. The responses of algae sizes to nutrients, especially silicate, total nitrogen and phosphate, were associated with the start and end of a life cycle. These correlations between size and nutrients were supported by both wavelet analysis and RDA. Moreover, the extremely high values at the end of the year were explained as algal recruitment from benthos. Our present study have introduced the worldwide scientific trends in phytoplankton studies using bibliometric analysis, demonstrated the temporal and spatial patterns of phytoplankton assemblages in response to environments within the downstream region of a large subtropical river in China. Our results will benefit the understanding of phytoplankton dynamics in freshwater ecosystems, as well as the large rivers all over the world