Tesi sul tema "Détection non supervisée d'anomalies"
Cita una fonte nei formati APA, MLA, Chicago, Harvard e in molti altri stili
Vedi i top-47 saggi (tesi di laurea o di dottorato) per l'attività di ricerca sul tema "Détection non supervisée d'anomalies".
Accanto a ogni fonte nell'elenco di riferimenti c'è un pulsante "Aggiungi alla bibliografia". Premilo e genereremo automaticamente la citazione bibliografica dell'opera scelta nello stile citazionale di cui hai bisogno: APA, MLA, Harvard, Chicago, Vancouver ecc.
Puoi anche scaricare il testo completo della pubblicazione scientifica nel formato .pdf e leggere online l'abstract (il sommario) dell'opera se è presente nei metadati.
Vedi le tesi di molte aree scientifiche e compila una bibliografia corretta.
Mazel, Johan. "Détection non supervisée d'anomalies dans les réseaux de communication". Phd thesis, INSA de Toulouse, 2011. http://tel.archives-ouvertes.fr/tel-00667654.
Jabiri, Fouad. "Applications de méthodes de classification non supervisées à la détection d'anomalies". Master's thesis, Université Laval, 2020. http://hdl.handle.net/20.500.11794/67914.
In this thesis, we will first present the binary tree partitioning algorithm and isolation forests. Binary trees are very popular classifiers in supervised machine learning. The isolation forest belongs to the family of unsupervised methods. It is an ensemble of binary trees used in common to isolate outlying instances. Subsequently, we will present the approach that we have named "Exponential smoothig" (or "pooling"). This technique consists in encoding sequences of variables of different lengths into a single vector of fixed size. Indeed, the objective of this thesis is to apply the algorithm of isolation forests to identify anomalies in insurance claim forms available in the database of a large Canadian insurance company in order to detect cases of fraud. However, a form is a sequence of claims. Each claim is characterized by a set of variables and thus it will be impossible to apply the isolation forest algorithm directly to this kind of data. It is for this reason that we are going to apply Exponential smoothing. Our application effectively isolates claims and abnormal forms, and we find that the latter tend to be audited by the company more often than regular forms.
Mazel, Johan. "Unsupervised network anomaly detection". Thesis, Toulouse, INSA, 2011. http://www.theses.fr/2011ISAT0024/document.
Anomaly detection has become a vital component of any network in today’s Internet. Ranging from non-malicious unexpected events such as flash-crowds and failures, to network attacks such as denials-of-service and network scans, network traffic anomalies can have serious detrimental effects on the performance and integrity of the network. The continuous arising of new anomalies and attacks create a continuous challenge to cope with events that put the network integrity at risk. Moreover, the inner polymorphic nature of traffic caused, among other things, by a highly changing protocol landscape, complicates anomaly detection system's task. In fact, most network anomaly detection systems proposed so far employ knowledge-dependent techniques, using either misuse detection signature-based detection methods or anomaly detection relying on supervised-learning techniques. However, both approaches present major limitations: the former fails to detect and characterize unknown anomalies (letting the network unprotected for long periods) and the latter requires training over labeled normal traffic, which is a difficult and expensive stage that need to be updated on a regular basis to follow network traffic evolution. Such limitations impose a serious bottleneck to the previously presented problem.We introduce an unsupervised approach to detect and characterize network anomalies, without relying on signatures, statistical training, or labeled traffic, which represents a significant step towards the autonomy of networks. Unsupervised detection is accomplished by means of robust data-clustering techniques, combining Sub-Space clustering with Evidence Accumulation or Inter-Clustering Results Association, to blindly identify anomalies in traffic flows. Correlating the results of several unsupervised detections is also performed to improve detection robustness. The correlation results are further used along other anomaly characteristics to build an anomaly hierarchy in terms of dangerousness. Characterization is then achieved by building efficient filtering rules to describe a detected anomaly. The detection and characterization performances and sensitivities to parameters are evaluated over a substantial subset of the MAWI repository which contains real network traffic traces.Our work shows that unsupervised learning techniques allow anomaly detection systems to isolate anomalous traffic without any previous knowledge. We think that this contribution constitutes a great step towards autonomous network anomaly detection.This PhD thesis has been funded through the ECODE project by the European Commission under the Framework Programme 7. The goal of this project is to develop, implement, and validate experimentally a cognitive routing system that meet the challenges experienced by the Internet in terms of manageability and security, availability and accountability, as well as routing system scalability and quality. The concerned use case inside the ECODE project is network anomaly
Cherdo, 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.
In 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
Barreyre, Clementine. "Statistiques en grande dimension pour la détection d'anomalies dans les données fonctionnelles issues des satellites". Thesis, Toulouse, INSA, 2018. http://www.theses.fr/2018ISAT0009/document.
In this PhD, we have developed statistical methods to detect abnormal events in all the functional data produced by the satellite all through its lifecycle. The data we are dealing with come from two main phases in the satellite’s life, telemetries and test data. A first work on this thesis was to understand how to highlight the outliers thanks to projections onto functional bases. On these projections, we have also applied several outlier detection methods, such as the One-Class SVM, the Local Outlier Factor (LOF). In addition to these two methods, we have developed our own outlier detection method, by taking into account the seasonality of the data we consider. Based on this study, we have developed an original procedure to select automatically the most interesting coefficients in a semi-supervised framework for the outlier detection, from a given projection. Our method is a multiple testing procedure where we apply the two sample-test to all the levels of coefficients.We have also chosen to analyze the covariance matrices representing the covariance of the te- lemetries between themselves for the outlier detection in multivariate data. In this purpose, we are comparing the covariance of a cluster of several telemetries deriving from two consecutive days, or consecutive orbit periods. We have applied three statistical tests targeting this same issue with different approaches. We have also developed an original asymptotic test, inspired by both first tests. In addition to the proof of the convergence of this test, we demonstrate thanks to examples that this new test is the most powerful. In this PhD, we have tackled several aspects of the anomaly detection in the functional data deriving from satellites. For each of these methods, we have detected all the major anomalies, improving significantly the false discovery rate
Boussik, Amine. "Apprentissage profond non-supervisé : Application à la détection de situations anormales dans l’environnement du train autonome". Electronic Thesis or Diss., Valenciennes, Université Polytechnique Hauts-de-France, 2023. http://www.theses.fr/2023UPHF0040.
The thesis addresses the challenges of monitoring the environment and detecting anomalies, especially obstacles, for an autonomous freight train. Although traditionally, rail transport was under human supervision, autonomous trains offer potential advantages in terms of costs, time, and safety. However, their operation in complex environments poses significant safety concerns. Instead of a supervised approach that requires costly and limited annotated data, this research adopts an unsupervised technique, using unlabeled data to detect anomalies based on methods capable of identifying atypical behaviors.Two environmental surveillance models are presented : the first, based on a convolutional autoencoder (CAE), is dedicated to identifying obstacles on the main track; the second, an advanced version incorporating the vision transformer (ViT), focuses on overall environmental surveillance. Both employ unsupervised learning techniques for anomaly detection.The results show that the highlighted method offers relevant insights for monitoring the environment of the autonomous freight train, holding potential to enhance its reliability and safety. The use of unsupervised techniques thus showcases the utility and relevance of their adoption in an application context for the autonomous train
Truong, Thi Bich Thanh. "Home Automation Monitoring for Assisted Living Services and Healthcare". Lorient, 2010. http://www.theses.fr/2010LORIS204.
With the development of technology and information, there are more and more opportunities and challenges for healthcare and assistance services for disabled people as well as the elderly. In this context, this PhD work proposes and demonstrates a new solution for home monitoring. Our approach is based on the idea that existing home automation and multimedia services provide some relevant information to be used as available sensors for remote monitoring. Through the analysis of user habits, our work includes two steps. In the first step, we automate a scenario identification, based on a combination of data mining, AI, graph theory and operational research algorithms to offer scenarios self adapting to user capabilities, while facilitating user access to the services. In the second step, this sensor information is used for alert management based on the anomaly detection, meaning a deviation of usual habits. These two steps provide a low level and non-intrusive personal monitoring while giving people more autonomy and confidence in their environments. A simulation model is developed in a first stage for the generation of user database without waiting for months monitoring user activities. This simulation data allows us to develop, tune and evaluate different aspects of our approach, before being applied in a real context. Then an experimentation through the IR recording is realized to monitor the user activities. The results of these real data allow us to evaluate the performance as well as the efficiency of our solution
Pantin, Jérémie. "Détection et caractérisation sémantique de données textuelles aberrantes". Electronic Thesis or Diss., Sorbonne université, 2023. https://accesdistant.sorbonne-universite.fr/login?url=https://theses-intra.sorbonne-universite.fr/2023SORUS347.pdf.
Machine learning answers to the problem of handling dedicated tasks with a wide variety of data. Such algorithms can be either simple or difficult to handle depending of the data. Low dimensional data (2-dimension or 3-dimension) with an intuitive representation (average of baguette price by years) are easier to interpret/explain for a human than data with thousands of dimensions. For low dimensional data, the error leads to a significant shift against normal data, but for the case of high dimensional data it is different. Outlier detection (or anomaly detection, or novelty detection) is the study of outlying observations for detecting what is normal and abnormal. Methods that perform such task are algorithms, methods or models that are based on data distributions. Different families of approaches can be found in the literature of outlier detection, and they are mainly independent of ground truth. They perform outlier analysis by detecting the principal behaviors of majority of observations. Thus, data that differ from normal distribution are considered noise or outlier. We detail the application of outlier detection with text. Despite recent progress in natural language processing, computer still lack profound understanding of human language in absence of information. For instance, the sentence "A smile is a curve that set everything straight" has several levels of understanding and a machine can encounter hardship to chose the right level of lecture. This thesis presents the analysis of high-dimensional outliers, applied to text. Recent advances in anomaly detection and outlier detection are not significantly represented with text data and we propose to highlight the main differences with high-dimensional outliers. We also approach ensemble methods that are nearly nonexistent in the literature for our context. Finally, an application of outlier detection for elevate results on abstractive summarization is conducted. We propose GenTO, a method that prepares and generates split of data in which anomalies and outliers are inserted. Based on this method, evaluation and benchmark of outlier detection approaches is proposed with documents. The proposed taxonomy allow to identify difficult and hierarchised outliers that the literature tackles without knowing. Also, learning without supervision often leads models to rely in some hyperparameter. For instance, Local Outlier Factor relies to the k-nearest neighbors for computing the local density. Thus, choosing the right value for k is crucial. In this regard, we explore the influence of such parameter for text data. While choosing one model can leads to obvious bias against real-world data, ensemble methods allow to mitigate such problem. They are particularly efficient with outlier analysis. Indeed, the selection of several values for one hyperparameter can help to detect strong outliers.Importance is then tackled and can help a human to understand the output of black box model. Thus, the interpretability of outlier detection models is questioned. We find that for numerous dataset, a low number of features can be selected as oracle. The association of complete models and restrained models helps to mitigate the black-box effect of some approaches. In some cases, outlier detection refers to noise removal or anomaly detection. Some applications can benefit from the characteristic of such task. Mail spam detection and fake news detection are one example, but we propose to use outlier detection approaches for weak signal exploration in marketing project. Thus, we find that the model of the literature help to improve unsupervised abstractive summarization, and also to find weak signals in text
Audibert, Julien. "Unsupervised anomaly detection in time-series". Electronic Thesis or Diss., Sorbonne université, 2021. http://www.theses.fr/2021SORUS358.
Anomaly detection in multivariate time series is a major issue in many fields. The increasing complexity of systems and the explosion of the amount of data have made its automation indispensable. This thesis proposes an unsupervised method for anomaly detection in multivariate time series called USAD. However, deep neural network methods suffer from a limitation in their ability to extract features from the data since they only rely on local information. To improve the performance of these methods, this thesis presents a feature engineering strategy that introduces non-local information. Finally, this thesis proposes a comparison of sixteen time series anomaly detection methods to understand whether the explosion in complexity of neural network methods proposed in the current literature is really necessary
Attal, Ferhat. "Classification de situations de conduite et détection des événements critiques d'un deux roues motorisé". Thesis, Paris Est, 2015. http://www.theses.fr/2015PEST1003/document.
This thesis aims to develop framework tools for analyzing and understanding the riding of Powered Two Wheelers (PTW). Experiments are conducted using instrumented PTW in real context including both normal (naturalistic) riding behaviors and critical riding behaviors (near fall and fall). The two objectives of this thesis are the riding patterns classification and critical riding events detection. In the first part of this thesis, a machine-learning framework is used for riding pattern recognition problem. Therefore, this problem is formulated as a classification task to identify the class of riding patterns. The approaches developed in this context have shown the interest to take into account the temporal aspect of the data in PTW riding. Moreover, we have shown the effectiveness of hidden Markov models for such problem. The second part of this thesis focuses on the development of the off-line detection and classification of critical riding events tools and the on-line fall detection. The problem of detection and classification of critical riding events has been performed towards two steps: (1) the segmentation step, where the multidimensional time of data were modeled and segmented by using a mixture model with quadratic logistic proportions; (2) the classification step, which consists in using a pattern recognition algorithm in order to assign each event by its extracted features to one of the three classes namely Fall, near Fall and Naturalistic riding. Regarding the fall detection problem, it is formulated as a sequential anomaly detection problem. The Multivariate CUmulative SUM (MCUSUM) control chart was applied on the data collected from sensors mounted on the motorcycle. The obtained results on a real database have shown the effectiveness of the proposed methodology for both riding pattern recognition and critical riding events detection problems
Putina, Andrian. "Unsupervised anomaly detection : methods and applications". Electronic Thesis or Diss., Institut polytechnique de Paris, 2022. http://www.theses.fr/2022IPPAT012.
An anomaly (also known as outlier) is an instance that significantly deviates from the rest of the input data and being defined by Hawkins as 'an observation, which deviates so much from other observations as to arouse suspicions that it was generated by a different mechanism'. Anomaly detection (also known as outlier or novelty detection) is thus the machine learning and data mining field with the purpose of identifying those instances whose features appear to be inconsistent with the remainder of the dataset. In many applications, correctly distinguishing the set of anomalous data points (outliers) from the set of normal ones (inliers) proves to be very important. A first application is data cleaning, i.e., identifying noisy and fallacious measurement in a dataset before further applying learning algorithms. However, with the explosive growth of data volume collectable from various sources, e.g., card transactions, internet connections, temperature measurements, etc. the use of anomaly detection becomes a crucial stand-alone task for continuous monitoring of the systems. In this context, anomaly detection can be used to detect ongoing intrusion attacks, faulty sensor networks or cancerous masses.The thesis proposes first a batch tree-based approach for unsupervised anomaly detection, called 'Random Histogram Forest (RHF)'. The algorithm solves the curse of dimensionality problem using the fourth central moment (aka kurtosis) in the model construction while boasting linear running time. A stream based anomaly detection engine, called 'ODS', that leverages DenStream, an unsupervised clustering technique is presented subsequently and finally Automated Anomaly Detection engine which alleviates the human effort required when dealing with several algorithm and hyper-parameters is presented as last contribution
Alaverdyan, Zaruhi. "Unsupervised representation learning for anomaly detection on neuroimaging. Application to epilepsy lesion detection on brain MRI". Thesis, Lyon, 2019. http://www.theses.fr/2019LYSEI005/document.
This work represents one attempt to develop a computer aided diagnosis system for epilepsy lesion detection based on neuroimaging data, in particular T1-weighted and FLAIR MR sequences. Given the complexity of the task and the lack of a representative voxel-level labeled data set, the adopted approach, first introduced in Azami et al., 2016, consists in casting the lesion detection task as a per-voxel outlier detection problem. The system is based on training a one-class SVM model for each voxel in the brain on a set of healthy controls, so as to model the normality of the voxel. The main focus of this work is to design representation learning mechanisms, capturing the most discriminant information from multimodality imaging. Manual features, designed to mimic the characteristics of certain epilepsy lesions, such as focal cortical dysplasia (FCD), on neuroimaging data, are tailored to individual pathologies and cannot discriminate a large range of epilepsy lesions. Such features reflect the known characteristics of lesion appearance; however, they might not be the most optimal ones for the task at hand. Our first contribution consists in proposing various unsupervised neural architectures as potential feature extracting mechanisms and, eventually, introducing a novel configuration of siamese networks, to be plugged into the outlier detection context. The proposed system, evaluated on a set of T1-weighted MRIs of epilepsy patients, showed a promising performance but a room for improvement as well. To this end, we considered extending the CAD system so as to accommodate multimodality data which offers complementary information on the problem at hand. Our second contribution, therefore, consists in proposing strategies to combine representations of different imaging modalities into a single framework for anomaly detection. The extended system showed a significant improvement on the task of epilepsy lesion detection on T1-weighted and FLAIR MR images. Our last contribution focuses on the integration of PET data into the system. Given the small number of available PET images, we make an attempt to synthesize PET data from the corresponding MRI acquisitions. Eventually we show an improved performance of the system when trained on the mixture of synthesized and real images
Ait, Saada Mira. "Unsupervised learning from textual data with neural text representations". Electronic Thesis or Diss., Université Paris Cité, 2023. http://www.theses.fr/2023UNIP7122.
The digital era generates enormous amounts of unstructured data such as images and documents, requiring specific processing methods to extract value from them. Textual data presents an additional challenge as it does not contain numerical values. Word embeddings are techniques that transform text into numerical data, enabling machine learning algorithms to process them. Unsupervised tasks are a major challenge in the industry as they allow value creation from large amounts of data without requiring costly manual labeling. In thesis we explore the use of Transformer models for unsupervised tasks such as clustering, anomaly detection, and data visualization. We also propose methodologies to better exploit multi-layer Transformer models in an unsupervised context to improve the quality and robustness of document clustering while avoiding the choice of which layer to use and the number of classes. Additionally, we investigate more deeply Transformer language models and their application to clustering, examining in particular transfer learning methods that involve fine-tuning pre-trained models on a different task to improve their quality for future tasks. We demonstrate through an empirical study that post-processing methods based on dimensionality reduction are more advantageous than fine-tuning strategies proposed in the literature. Finally, we propose a framework for detecting text anomalies in French adapted to two cases: one where the data concerns a specific topic and the other where the data has multiple sub-topics. In both cases, we obtain superior results to the state of the art with significantly lower computation time
Huck, Alexis. "Analyse non-supervisée d’images hyperspectrales : démixage linéaire et détection d’anomalies". Aix-Marseille 3, 2009. http://www.theses.fr/2009AIX30036.
This thesis focusses on two research fields regarding unsupervised analysis of hyperspectral images (HSIs). Under the assumptions of the linear spectral mixing model, the formalism of Non-Negative Matrix Factorization is investigated for unmixing purposes. We propose judicious spectral and spatial a priori knowledge to regularize the problem. In addition, we propose an estimator for the projected gradient optimal step-size. Thus, suitably regularized NMF is shown to be a relevant approach to unmix HSIs. Then, the problem of anomaly detection is considered. We propose an algorithm for Anomalous Component Pursuit (ACP), simultaneously based on projection pursuit and on a probabilistic model and hypothesis testing. ACP detects the anomalies with a constant false alarm rate and discriminates them into spectrally homogeneous classes
Benammar, Riyadh. "Détection non-supervisée de motifs dans les partitions musicales manuscrites". Thesis, Lyon, 2019. http://www.theses.fr/2019LYSEI112.
This thesis is part of the data mining applied to ancient handwritten music scores and aims at a search for frequent melodic or rhythmic motifs defined as repetitive note sequences with characteristic properties. There are a large number of possible variations of motifs: transpositions, inversions and so-called "mirror" motifs. These motifs allow musicologists to have a level of in-depth analysis on the works of a composer or a musical style. In a context of exploring large corpora where scores are just digitized and not transcribed, an automated search for motifs that verify targeted constraints becomes an essential tool for their study. To achieve the objective of detecting frequent motifs without prior knowledge, we started from images of digitized scores. After pre-processing steps on the image, we exploited and adapted a model for detecting and recognizing musical primitives (note-heads, stems...) from the family of Region-Proposal CNN (RPN) convolution neural networks. We then developed a primitive encoding method to generate a sequence of notes without the complex task of transcribing the entire manuscript work. This sequence was then analyzed using the CSMA (Constraint String Mining Algorithm) approach designed to detect the frequent motifs present in one or more sequences, taking into account constraints on their frequency and length, as well as the size and number of gaps allowed within the motifs. The gap was then studied to avoid recognition errors produced by the RPN network, thus avoiding the implementation of a post-correction system for transcription errors. The work was finally validated by the study of musical motifs for composers identification and classification
Lung-Yut-Fong, Alexandre. "Détection robuste de ruptures pour les signaux multidimensionnels : application à la détection d'anomalies dans les réseaux". Paris, Télécom ParisTech, 2011. https://pastel.hal.science/pastel-00675543.
The aim of this work is to propose non-parametric change-point detection methods. The main application of such methods is the use of data recorded by a collection of network sensors to detect malevolent attacks. The first contribution of the thesis work is a decentralized anomaly detector. Each network sensor applies a rank-based change-point detection test, and the final decision is taken by a fusion center which aggregates the information transmitted by the sensors. This method is able to process a huge amount of data, thanks to a clever filtering step. In the second contribution, we take into account the dependencies between the different sensors to improve the detection performance. Based on homogeneity tests that we have proposed to assess the similarity between different sets of data, the robust detection methods that we have designed are able to find one or more change-point in a multidimensional signal. We thus obtained robust and versatile methods, with strong theoretical properties, to solve a large collection of segmentation problems: network anomaly detection, econometrics, DNA analysis for cancer prognosis… The methods that we proposed are particularly adequate when the characteristics of the analyzed data are unknown
Laurendin, Olivier. "Identification de situations de mise en danger d'usagers aux alentours de portes automatiques de train par détection d'anomalies auto-supervisée". Electronic Thesis or Diss., Ecole nationale supérieure Mines-Télécom Lille Douai, 2023. http://www.theses.fr/2023MTLD0009.
This work is part of a research project to develop an autonomous passenger train operating on French regional lines. Ultimately, the developed autonomous train prototype aims for the maximum degree of railway automation, for which all train operating functions currently under the responsability of on-board staff will be handled by technical systems. We are interested here in the automation of one of these functions, namely the safe operation of the train’s external doors.Numerous dangerous situations for passagers during door closing have been identified, and some are now among the main sources of passager incidents. The aim of this thesis is to develop a detection system to identify these events based on the automatic analysis of on-board surveillance system video streams.This thesis addresses this problem by applying deep neural networks to detect, locate and identify any dangerous event related to pedestrians and doors present in video streams provided by fish-eye cameras installed on the ceiling of the train boarding platforms.As instances of these dangerous events occur very rarely during train operation, the proposed solution is based on the notion of anomaly, defined as an unknown or unexpected event in a given context of « normality ». The proposed neural architecture therefore constitutes a model of normality and identifies as abnormal any aberrant data that deviates from it.The neural network architecture we are proposing breaks down into two specialized branches capable of learning a model of normal interaction between two objects. Each branch is trained using learned proxy tasks capable of modelling different aspects of normality assumed to be relevant for the detection of anomalies associated with pedestrians and doors.As no image dataset related to our use case exists in the literature, we have collected and annotated a set of image sequences for training and evaluating our architectures. These sequences depict users in the vicinity of a train doors replica and in a real train instrumented as part of the research project. Finaly, we have developed metrics for evaluating the effectiveness of our models in order to test their operational applicability
Peng, Anrong. "Segmentation statistique non supervisée d'images et de détection de contours par filtrage". Compiègne, 1992. http://www.theses.fr/1992COMPD512.
Lung-Yut-Fong, Alexandre. "Détection de ruptures pour les signaux multidimensionnels. Application à la détection d'anomalies dans les réseaux". Phd thesis, Télécom ParisTech, 2011. http://pastel.archives-ouvertes.fr/pastel-00675543.
Dubois, Rémi. "Application des nouvelles méthodes d'apprentissage à la détection précoce d'anomalies en électrocardiographie". Paris 6, 2004. https://pastel.archives-ouvertes.fr/pastel-00000571.
Dubois, R. "Application des nouvelles méthodes d'apprentissage à la détection précoce d'anomalies cardiaques en électrocardiographie". Phd thesis, Université Pierre et Marie Curie - Paris VI, 2004. http://pastel.archives-ouvertes.fr/pastel-00000571.
Nasser, Alissar. "Contribution à la classification non supervisée par machines à noyaux". Littoral, 2007. http://www.theses.fr/2007DUNK0182.
Unsupervised classification has emerged as a popular technique for pattern recognition, image processing, and data mining. This is due to the development of advanced data measurements tools and data storage devices resulting in a huge quantity of data. This makes it necessary to analyze these data in order to extract some useful information. Unsupervised classification is one of the well-studied techniques, which concerns the partitioning of similar objects into clusters without any prior knowledge, such that objects in the same cluster share some unique properties. Two main categories of methods exist : (1) clustering methods in the multidimensional space and (2) projection methods for exploratory data analysis. The first category seeks zones/groups of high densities whereas the second category provides an accurate image on the plane of the multidimensional data. One of convenient lethods is by combining these two categories together in a way that involves a human operator into the process of structure analysis. Recently, Kernel machines gained a success in unsupervised classification. Instead, of projecting or classifying data directly in their input space, one transforms it into a high dimensional space called feature space and then applies any traditional projection technique such as Principal Components Analysis (PCA) or any clustering method such as K-means algorithm. The logic behind kernel is to enhance those features of the input data which make distinct pattern classes separate from each other. The present thesis shows the contribution of kernel machines in unsupervised classification, particularly in projection and classification methods. It first presents traditional projection methods and then present kernel Principal Components Analysis (kPCA). Spectral classification and kernel K-means clustering algortihm. The problems of adjusting kernel parameters and estimating the number of classes are studied. More over samples on synthetic and real data are executed ; results from various presented methods are compared. These clustering approaches are finally applied for the assistance to the detection of audio events in public transport
Luvison, Bertrand. "Détection non supervisée d'évènements rares dans un flot vidéo : application à la surveillance d'espaces publics". Phd thesis, Université Blaise Pascal - Clermont-Ferrand II, 2010. http://tel.archives-ouvertes.fr/tel-00626490.
Laby, Romain. "Détection et localisation d'anomalies dans des données hétérogènes en utilisant des modèles graphiques non orientés mixtes". Electronic Thesis or Diss., Paris, ENST, 2017. http://www.theses.fr/2017ENST0026.
This thesis revolves around an industrial need of Thales Système Aéroportés and the RBE2 combat radar equipping Dassault Rafale fighter aircraft. It develops a methodology for locating anomalies in heterogeneous data stream using a mixed, non-orientation and peer-to-peer graphical model. The data are a mixture of categorical and quantitative variables, and the model is learned from a data set that is assumed not to contain abnormal data. Anomaly localization algorithms use an adapted version of the CUSUM algorithm, whose decision function is based on the calculation of conditional likelihood ratios. This function allows the detection of variable anomalies per variable and the precise localization of the variables involved in the anomaly
Goubet, Étienne. "Contrôle non destructif par analyse supervisée d'images 3D ultrasonores". Cachan, Ecole normale supérieure, 1999. http://www.theses.fr/1999DENS0011.
El, Khoury Elie. "Indexation vidéo non-supervisée basée sur la caractérisation des personnes". Phd thesis, Université Paul Sabatier - Toulouse III, 2010. http://tel.archives-ouvertes.fr/tel-00515424.
Labadié, Alexandre. "Segmentation thématique de texte linéaire et non-supervisée : détection active et passive des frontières thématiques en Français". Phd thesis, Montpellier 2, 2008. http://www.theses.fr/2008MON20159.
Labadié, Alexandre. "Segmentation thématique de texte linéaire et non-supervisée : détection active et passive des frontières thématiques en Français". Phd thesis, Université Montpellier II - Sciences et Techniques du Languedoc, 2008. http://tel.archives-ouvertes.fr/tel-00364848.
Gan, Changquan. "Une approche de classification non supervisée basée sur la notion des K plus proches voisins". Compiègne, 1994. http://www.theses.fr/1994COMP765S.
Thépaut, Solène. "Problèmes de clustering liés à la synchronie en écologie : estimation de rang effectif et détection de ruptures sur les arbres". Thesis, Université Paris-Saclay (ComUE), 2019. http://www.theses.fr/2019SACLS477/document.
In the view of actual global changes widely caused by human activities, it becomes urgent to understand the drivers of communities' stability. Synchrony between time series of abundances is one of the most important mechanisms. This thesis offers three different angles in order to answer different questions linked to interspecific and spatial synchrony. The works presented find applications beyond the ecological frame. A first chapter is dedicated to the estimation of effective rank of matrices in ℝ or ℂ. We offer tools allowing to measure the synchronisation rate of observations matrices. In the second chapter, we base on the existing work on change-points detection problem on chains in order to offer algorithms which detects change-points on trees. The methods can be used with most data that have to be represented as a tree. In order to study the link between interspecific synchrony and long term tendencies or traits of butterflies species, we offer in the last chapter adaptation of clustering and supervised machine learning methods, such as Random Forest or Artificial Neural Networks to ecological data
Goulet, Sylvain. "Techniques d'identification d'entités nommées et de classification non-supervisée pour des requêtes de recherche web à l'aide d'informations contenues dans les pages web visitées". Mémoire, Université de Sherbrooke, 2014. http://hdl.handle.net/11143/5387.
Varejão, Andreão Rodrigo. "Segmentation de battements ECG par approche markovienne : application à la détection d'ischémies". Evry, Institut national des télécommunications, 2004. http://www.theses.fr/2004TELE0004.
Ambulatory electrocardiography (AECG) provides precise and rich information from the clinical point of view for the diagnostic of cardiac diseases and particularly myocardial ischemia in patients with coronary disease. Early detection of myocardial ischemia allows fast diagnostic and makes treatment more effective. Ischemic episodes are detected through the ST-segment deviation function, wich is built after analysis of each heartbeat. In this context, we propose a system combining a Markovian approach and a heuristic approach to perform automatic ischemic episode detection. Our markovian approach extracts from the ECG signal the information needed to perform ST-sergment deviation analysis. It is able to take into account complex morphologies thanks to the use of individual HMM to model each beat waveform (P, QRS and T). In addition, our original non supervised training strategy provides HMM parameter adaptation to the ECG signal of each patient. To classify the ECG signal in terms of a specific abnormality, we added a set of rules to manage the information extracted from the signal. We also explored the information fusion obtained from different leads yielding to more reliable detection results. Finally, we assessed our system performance over two AECG databases. Different problems were concerned QRS complex detection, waveform sergmentation precision, premature ventricular contraction beat detection and ischemic episode detection. All results attest the interest in the approach proposed and compare favourably to the state of the art
Thivin, Solenne. "Détection automatique de cibles dans des fonds complexes. Pour des images ou séquences d'images". Thesis, Université Paris-Saclay (ComUE), 2015. http://www.theses.fr/2015SACLS235/document.
During this PHD, we developped an detection algorithm. Our principal objective was to detect small targets in a complex background like clouds for example.For this, we used the spatial covariate structure of the real images.First, we developped a collection of models for this covariate structure. Then, we selected a special model in the previous collection. Once the model selected, we applied the likelihood ratio test to detect the potential targets.We finally studied the performances of our algorithm by testing it on simulated and real images
Kassab, Randa. "Analyse des propriétés stationnaires et des propriétés émergentes dans les flux d'information changeant au cours du temps". Thesis, Nancy 1, 2009. http://www.theses.fr/2009NAN10027/document.
Many applications produce and receive continuous, unlimited, and high-speed data streams. This raises obvious problems of storage, treatment and analysis of data, which are only just beginning to be treated in the domain of data streams. On the one hand, it is a question of treating data streams on the fly without having to memorize all the data. On the other hand, it is also a question of analyzing, in a simultaneous and concurrent manner, the regularities inherent in the data stream as well as the novelties, exceptions, or changes occurring in this stream over time. The main contribution of this thesis concerns the development of a new machine learning approach - called ILoNDF - which is based on novelty detection principle. The learning of this model is, contrary to that of its former self, driven not only by the novelty part in the input data but also by the data itself. Thereby, ILoNDF can continuously extract new knowledge relating to the relative frequencies of the data and their variables. This makes it more robust against noise. Being operated in an on-line mode without repeated training, ILoNDF can further address the primary challenges for managing data streams. Firstly, we focus on the study of ILoNDF's behavior for one-class classification when dealing with high-dimensional noisy data. This study enabled us to highlight the pure learning capacities of ILoNDF with respect to the key classification methods suggested until now. Next, we are particularly involved in the adaptation of ILoNDF to the specific context of information filtering. Our goal is to set up user-oriented filtering strategies rather than system-oriented in following two types of directions. The first direction concerns user modeling relying on the model ILoNDF. This provides a new way of looking at user's need in terms of specificity, exhaustivity and contradictory profile-contributing criteria. These criteria go on to estimate the relative importance the user might attach to precision and recall. The filtering threshold can then be adjusted taking into account this knowledge about user's need. The second direction, complementary to the first one, concerns the refinement of ILoNDF's functionality in order to confer it the capacity of tracking drifting user's need over time. Finally, we consider the generalization of our previous work to the case where streaming data can be divided into multiple classes
Le, Gorrec Luce. "Équilibrage bi-stochastique des matrices pour la détection de structures par blocs et applications". Thesis, Toulouse 3, 2019. http://www.theses.fr/2019TOU30136.
The detection of block structures in matrices is an important challenge. First in data analysis where matrices are a key tool for data representation, as data tables or adjacency matrices. Indeed, for the first one, finding a co-clustering is equivalent to finding a row and column block structure of the matrix. For the second one, finding a structure of diagonal dominant blocks leads to a clustering of the data. Moreover, block structure detection is also usefull for the resolution of linear systems. For instance, it helps to create efficient Block Jacobi precoditioners or to find groups of rows that are strongly decorrelated in order to apply a solver such as Block Cimmino. In this dissertation, we focus our analysis on the detection of dominant diagonal block structures by symmetrically permuting the rows and columns of matrices. Lots of algorithms have been designed that aim to highlight such structures. Among them, spectral algorithms play a key role. They can be divided into two kinds. The first one consists of algorithms that first project the matrix rows onto a low-dimensional space generated by the matrix leading eigenvectors, and then apply a procedure such as a k-means on the reduced data. Their main drawbacks is that the knowledge of number of clusters to uncover is required. The second kind consists of iterative procedures that look for the k-th best partition into two subblocks of the matrix at step k. However, if the matrix structure shows more than two blocks, the best partition into two blocks may be a poor fit to the matrix groundtruth structure. Hence, we propose a spectral algorithm that deals with both issues described above. To that end, we preprocess the matrix with a doubly-stochastic scaling, which leverages the blocks. First we show the benefits of using such a scaling by using it as a preprocessing for the Louvain's algorithm, in order to uncover community structures in networks. We also investigate several global modularity measures designed for quantifying the consistency of a block structure. We generalise them to make them able to handle doubly-stochastic matrices, and thus we remark that our scaling tends to unify these measures. Then, we describe our algorithm that is based on spectral elements of the scaled matrix. Our method is built on the principle that leading singular vectors of a doubly-stochastic matrix should have a staircase pattern when their coordinates are sorted in the increasing order, under the condition that the matrix shows a hidden block structure. Tools from signal processing-that have been initially designed to detect jumps in signals-are applied to the sorted vectors in order to detect steps in these vectors, and thus to find the separations between the blocks. However, these tools are not specifically designed to this purpose. Hence procedures that we have implemented to answer the encountered issues are also described. We then propose three applications for the matrices block structure detection. First, community detection in networks, and the design of efficient Block Jacobi type preconditioners for solving linear systems. For these applications, we compare the results of our algorithm with those of algorithms that have been designed on purpose. Finally, we deal with the dialogue act detection in a discorsre, using the STAC database that consists in a chat of online players of " The Settlers of Catan ". To that end we connect classical clustering algorithms with a BiLSTM neural network taht preprocesses the dialogue unities. Finally, we conclude by giving some preliminary remarks about the extension of our method to rectangular matrices
Khichane, Abderaouf. "Diagnostic of performance by data interpretation for 5G cloud native network functions". Electronic Thesis or Diss., université Paris-Saclay, 2024. http://www.theses.fr/2024UPASG017.
Operators today are facing a profound and inevitable evolution of services and infrastructure. They are constantly pressured to accelerate the renewal of their offerings to meet new challenges and opportunities. It is in this context that the concept of "Cloud-native" network functions [1][2][3] is gaining increasing significance. Drawing inspiration from the IT world where "Cloud readiness" has already proven its worth, the idea of cloudifying network functions involves implementing scalable and self-healing functions while providing generic APIs accessible through their management and orchestration systems. However, the transition to a "Cloud-native" model is not limited to encapsulating network functions in virtual machines. It requires an adaptation, even a total redesign, of network functions.In this context, microservices architectures [4] become essential for the design of cloud-native 5G applications. Decomposing applications into independent services brings flexibility in terms of i) development, ii) deployment, and iii) scalability. Nevertheless, adopting this new architectural paradigm for virtualized network functions raises new questions about orchestration and automation operations. In particular, observability represents a cornerstone in monitoring 5G functions to provide the highest level of customer satisfaction. This functionality involves activities related to measuring, collecting, and analyzing telemetry data from both the operator's infrastructure and the applications running on it. Observability enables a deep understanding of network behavior and the anticipation of service quality degradation. Various observability approaches are proposed in the literature [5], allowing the analysis of the behavior of cloud-native IT applications and the implementation of necessary remediation actions.In this context, telemetry data provides precise information about the state of operator networks. However, the complexity of the operator's software-defined infrastructure and the volume of data [6] to be processed require the development of new techniques capable of detecting real-time risk situations and making the right decisions, for example, to avoid a violation of the Service Level Agreement (SLA). This is the framework in which the work of this thesis is situated
Kessi, Louisa. "Unsupervised detection based on spatial relationships : Application for object detection and recognition of colored business document structures". Thesis, Lyon, 2018. http://www.theses.fr/2018LYSEI068.
This digital revolution introduces new services and new usages in numerous domains. The advent of the digitization of documents and the automatization of their processing constitutes a great cultural and economic revolution. In this context, computer vision provides numerous applications and impacts our daily lives and businesses. Behind computer-vision technology, fundamental concepts, methodologies, and algorithms have been developed worldwide in the last fifty years. Today, computer vision technologies arrive to maturity and become a reality in many domains. Computer-vision systems reach high performance thanks to the large amount of data and the increasing performance of the hardware. Despite the success of computer-vision applications, however, numerous other applications require more research, new methodologies, and novel algorithms. Among the difficult problems encountered in the computer-vision domain, detection remains a challenging task. Detection consists of localizing and recognizing an object in an image. This problem is far more difficult than the problem of recognition alone. Among the numerous applications based on detection, object detection in a natural scene is the most popular application in the computer-vision community. This work is about the detection tasks and its applications
Kassab, Randa. "Analyse des propriétés stationnaires et des propriétés émergentes dans les flux d'informations changeant au cours du temps". Phd thesis, Université Henri Poincaré - Nancy I, 2009. http://tel.archives-ouvertes.fr/tel-00402644.
L'apport de ce travail de thèse réside principalement dans le développement d'un modèle d'apprentissage - nommé ILoNDF - fondé sur le principe de la détection de nouveauté. L'apprentissage de ce modèle est, contrairement à sa version de départ, guidé non seulement par la nouveauté qu'apporte une donnée d'entrée mais également par la donnée elle-même. De ce fait, le modèle ILoNDF peut acquérir constamment de nouvelles connaissances relatives aux fréquences d'occurrence des données et de leurs variables, ce qui le rend moins sensible au bruit. De plus, doté d'un fonctionnement en ligne sans répétition d'apprentissage, ce modèle répond aux exigences les plus fortes liées au traitement des flux de données.
Dans un premier temps, notre travail se focalise sur l'étude du comportement du modèle ILoNDF dans le cadre général de la classification à partir d'une seule classe en partant de l'exploitation des données fortement multidimensionnelles et bruitées. Ce type d'étude nous a permis de mettre en évidence les capacités d'apprentissage pures du modèle ILoNDF vis-à-vis de l'ensemble des méthodes proposées jusqu'à présent. Dans un deuxième temps, nous nous intéressons plus particulièrement à l'adaptation fine du modèle au cadre précis du filtrage d'informations. Notre objectif est de mettre en place une stratégie de filtrage orientée-utilisateur plutôt qu'orientée-système, et ceci notamment en suivant deux types de directions. La première direction concerne la modélisation utilisateur à l'aide du modèle ILoNDF. Cette modélisation fournit une nouvelle manière de regarder le profil utilisateur en termes de critères de spécificité, d'exhaustivité et de contradiction. Ceci permet, entre autres, d'optimiser le seuil de filtrage en tenant compte de l'importance que pourrait donner l'utilisateur à la précision et au rappel. La seconde direction, complémentaire de la première, concerne le raffinement des fonctionnalités du modèle ILoNDF en le dotant d'une capacité à s'adapter à la dérive du besoin de l'utilisateur au cours du temps. Enfin, nous nous attachons à la généralisation de notre travail antérieur au cas où les données arrivant en flux peuvent être réparties en classes multiples.
Soltani, Mariem. "Partitionnement des images hyperspectrales de grande dimension spatiale par propagation d'affinité". Thesis, Rennes 1, 2014. http://www.theses.fr/2014REN1S099/document.
The interest in hyperspectral image data has been constantly increasing during the last years. Indeed, hyperspectral images provide more detailed information about the spectral properties of a scene and allow a more precise discrimination of objects than traditional color images or even multispectral images. High spatial and spectral resolutions of hyperspectral images enable to precisely characterize the information pixel content. Though the potentialities of hyperspectral technology appear to be relatively wide, the analysis and the treatment of these data remain complex. In fact, exploiting such large data sets presents a great challenge. In this thesis, we are mainly interested in the reduction and partitioning of hyperspectral images of high spatial dimension. The proposed approach consists essentially of two steps: features extraction and classification of pixels of an image. A new approach for features extraction based on spatial and spectral tri-occurrences matrices defined on cubic neighborhoods is proposed. A comparative study shows the discrimination power of these new features over conventional ones as well as spectral signatures. Concerning the classification step, we are mainly interested in this thesis to the unsupervised and non-parametric classification approach because it has several advantages: no a priori knowledge, image partitioning for any application domain, and adaptability to the image information content. A comparative study of the most well-known semi-supervised (knowledge of number of classes) and unsupervised non-parametric methods (K-means, FCM, ISODATA, AP) showed the superiority of affinity propagation (AP). Despite its high correct classification rate, affinity propagation has two major drawbacks. Firstly, the number of classes is over-estimated when the preference parameter p value is initialized as the median value of the similarity matrix. Secondly, the partitioning of large size hyperspectral images is hampered by its quadratic computational complexity. Therefore, its application to this data type remains impossible. To overcome these two drawbacks, we propose an approach which consists of reducing the number of pixels to be classified before the application of AP by automatically grouping data points with high similarity. We also introduce a step to optimize the preference parameter value by maximizing a criterion related to the interclass variance, in order to correctly estimate the number of classes. The proposed approach was successfully applied on synthetic images, mono-component and multi-component and showed a consistent discrimination of obtained classes. It was also successfully applied and compared on hyperspectral images of high spatial dimension (1000 × 1000 pixels × 62 bands) in the context of a real application for the detection of invasive and non-invasive vegetation species
Musé, Pablo. "Sur la définition et la reconnaissance des formes planes dans les images numériques". Phd thesis, École normale supérieure de Cachan - ENS Cachan, 2004. http://tel.archives-ouvertes.fr/tel-00133648.
Ternynck, Camille. "Contributions à la modélisation de données spatiales et fonctionnelles : applications". Thesis, Lille 3, 2014. http://www.theses.fr/2014LIL30062/document.
In this dissertation, we are interested in nonparametric modeling of spatial and/or functional data, more specifically based on kernel method. Generally, the samples we have considered for establishing asymptotic properties of the proposed estimators are constituted of dependent variables. The specificity of the studied methods lies in the fact that the estimators take into account the structure of the dependence of the considered data.In a first part, we study real variables spatially dependent. We propose a new kernel approach to estimating spatial probability density of the mode and regression functions. The distinctive feature of this approach is that it allows taking into account both the proximity between observations and that between sites. We study the asymptotic behaviors of the proposed estimates as well as their applications to simulated and real data. In a second part, we are interested in modeling data valued in a space of infinite dimension or so-called "functional data". As a first step, we adapt the nonparametric regression model, introduced in the first part, to spatially functional dependent data framework. We get convergence results as well as numerical results. Then, later, we study time series regression model in which explanatory variables are functional and the innovation process is autoregressive. We propose a procedure which allows us to take into account information contained in the error process. After showing asymptotic behavior of the proposed kernel estimate, we study its performance on simulated and real data.The third part is devoted to applications. First of all, we present unsupervised classificationresults of simulated and real spatial data (multivariate). The considered classification method is based on the estimation of spatial mode, obtained from the spatial density function introduced in the first part of this thesis. Then, we apply this classification method based on the mode as well as other unsupervised classification methods of the literature on hydrological data of functional nature. Lastly, this classification of hydrological data has led us to apply change point detection tools on these functional data
Frévent, Camille. "Contribution to spatial statistics for high-dimensional and survival data". Electronic Thesis or Diss., Université de Lille (2022-....), 2022. http://www.theses.fr/2022ULILS032.
In this thesis, we are interested in statistical spatial learning for high-dimensional and survival data. The objective is to develop unsupervised cluster detection methods by means of spatial scan statistics in the contexts of functional data analysis in one hand and survival data analysis in the other hand. In the first two chapters, we consider univariate and multivariate functional data measured spatially in a geographical area. We propose both parametric and nonparametric spatial scan statistics in this framework. These univariate and multivariate functional approaches avoid the loss of information respectively of a univariate method or a multivariate method applied on the average of the observations during the study period. We study the new methods' performances in simulation studies before applying them on economic and environmental real data. We are also interested in spatial cluster detection of survival data. Although there exist already spatial scan statistics approaches in this framework in the literature, these do not take into account a potential correlation of survival times between individuals of the same spatial unit. Moreover, the spatial nature of the data implies a potential dependence between the spatial units, which should be taken into account. The originality of our proposed method is to introduce a spatial scan statistic based on a Cox model with a spatial frailty, allowing to take into account both the potential correlation between the survival times of the individuals of the same spatial unit and the potential dependence between the spatial units. We compare the performances of this new approach with the existing methods and apply them on real data corresponding to survival times of elderly people with end-stage kidney failure in northern France. Finally, we propose a number of perspectives to our work, both in a direct extension of this thesis in the framework of spatial scan statistics for high-dimensional and survival data, but also perspectives in a broader context of unsupervised spatial analysis (spatial clustering for high-dimensional data (tensors)), and supervised spatial learning (regression)
Zhao, Zilong. "Extracting knowledge from macroeconomic data, images and unreliable data". Thesis, Université Grenoble Alpes, 2020. http://www.theses.fr/2020GRALT074.
System identification and machine learning are two similar concepts independently used in automatic and computer science community. System identification uses statistical methods to build mathematical models of dynamical systems from measured data. Machine learning algorithms build a mathematical model based on sample data, known as "training data" (clean or not), in order to make predictions or decisions without being explicitly programmed to do so. Except prediction accuracy, converging speed and stability are another two key factors to evaluate the training process, especially in the online learning scenario, and these properties have already been well studied in control theory. Therefore, this thesis will implement the interdisciplinary researches for following topic: 1) System identification and optimal control on macroeconomic data: We first modelize the China macroeconomic data on Vector Auto-Regression (VAR) model, then identify the cointegration relation between variables and use Vector Error Correction Model (VECM) to study the short-time fluctuations around the long-term equilibrium, Granger Causality is also studied with VECM. This work reveals the trend of China's economic growth transition: from export-oriented to consumption-oriented; Due to limitation of China economic data, we turn to use France macroeconomic data in the second study. We represent the model in state-space, put the model into a feedback control framework, the controller is designed by Linear-Quadratic Regulator (LQR). The system can apply the control law to bring the system to a desired state. We can also impose perturbations on outputs and constraints on inputs, which emulates the real-world situation of economic crisis. Economists can observe the recovery trajectory of economy, which gives meaningful implications for policy-making. 2) Using control theory to improve the online learning of deep neural network: We propose a performance-based learning rate algorithm: E (Exponential)/PD (Proportional Derivative) feedback control, which consider the Convolutional Neural Network (CNN) as plant, learning rate as control signal and loss value as error signal. Results show that E/PD outperforms the state-of-the-art in final accuracy, final loss and converging speed, and the result are also more stable. However, one observation from E/PD experiments is that learning rate decreases while loss continuously decreases. But loss decreases mean model approaches optimum, we should not decrease the learning rate. To prevent this, we propose an event-based E/PD. Results show that it improves E/PD in final accuracy, final loss and converging speed; Another observation from E/PD experiment is that online learning fixes a constant training epoch for each batch. Since E/PD converges fast, the significant improvement only comes from the beginning epochs. Therefore, we propose another event-based E/PD, which inspects the historical loss, when the progress of training is lower than a certain threshold, we turn to next batch. Results show that it can save up to 67% epochs on CIFAR-10 dataset without degrading much performance. 3) Machine learning out of unreliable data: We propose a generic framework: Robust Anomaly Detector (RAD), The data selection part of RAD is a two-layer framework, where the first layer is used to filter out the suspicious data, and the second layer detects the anomaly patterns from the remaining data. We also derive three variations of RAD namely, voting, active learning and slim, which use additional information, e.g., opinions of conflicting classifiers and queries of oracles. We iteratively update the historical selected data to improve accumulated data quality. Results show that RAD can continuously improve model's performance under the presence of noise on labels. Three variations of RAD show they can all improve the original setting, and the RAD Active Learning performs almost as good as the case where there is no noise on labels
Puigt, Matthieu. "Méthodes de séparation aveugle de sources fondées sur des transformées temps-fréquence : application à des signaux de parole". Toulouse 3, 2007. http://thesesups.ups-tlse.fr/217/.
Several time-frequency (TF) blind source separation (BSS) methods have been proposed in this thesis. In the systems output that have been used, a contribution of each source is estimated, using only mixed signals. All the methods proposed in this manuscript find tiny TF zones where only one source is active and estimate the mixing parameters in these zones. These approaches are particularly well suited for non-stationary sources (speech, music). We first studied and improved linear instantaneous methods based on variance or correlation criteria, that have been previously proposed by our team. They yield excellent performance for signals of speech and can also separate spectra from astrophysical data. However, the nature of the mixtures that they can process limits their application fields. We have extended these approaches to more realistic mixtures. The first extensions consider attenuated and delayed mixtures of sources, which corresponds to mixtures in anechoic chamber. They require less restrictive sparsity assumptions than some approaches previously proposed in the literature, while addressing the same type of mixtures. We have studied the contribution of clustering techniques to our approaches and have achieved good performance for mixtures of speech signals. Lastly, a theoretical extension of these methods to general convolutive mixtures is described. It needs strong sparsity hypotheses and we have to solve classical indeterminacies of frequency-domain BSS methods
Kalinicheva, Ekaterina. "Unsupervised satellite image time series analysis using deep learning techniques". Electronic Thesis or Diss., Sorbonne université, 2020. http://www.theses.fr/2020SORUS335.
This thesis presents a set of unsupervised algorithms for satellite image time series (SITS) analysis. Our methods exploit machine learning algorithms and, in particular, neural networks to detect different spatio-temporal entities and their eventual changes in the time.In our thesis, we aim to identify three different types of temporal behavior: no change areas, seasonal changes (vegetation and other phenomena that have seasonal recurrence) and non-trivial changes (permanent changes such as constructions or demolishment, crop rotation, etc). Therefore, we propose two frameworks: one for detection and clustering of non-trivial changes and another for clustering of “stable” areas (seasonal changes and no change areas). The first framework is composed of two steps which are bi-temporal change detection and the interpretation of detected changes in a multi-temporal context with graph-based approaches. The bi-temporal change detection is performed for each pair of consecutive images of the SITS and is based on feature translation with autoencoders (AEs). At the next step, the changes from different timestamps that belong to the same geographic area form evolution change graphs. The graphs are then clustered using a recurrent neural networks AE model to identify different types of change behavior. For the second framework, we propose an approach for object-based SITS clustering. First, we encode SITS with a multi-view 3D convolutional AE in a single image. Second, we perform a two steps SITS segmentation using the encoded SITS and original images. Finally, the obtained segments are clustered exploiting their encoded descriptors
Tonnelier, Emeric. "Apprentissage de représentations pour les traces de mobilité". Electronic Thesis or Diss., Sorbonne université, 2019. http://www.theses.fr/2019SORUS389.
Urban transport is a crucial issue for territories management. In large cities, many inhabitants have to rely on urban public transport to move around, go to work, visit friends. Historically, urban transportation analysis is based on surveys. Questions are ask to a panel of users, leading to the introduction of various bias and no dynamic informations. Since the late 1990s, we see the emergence of new types of data (GPS, smart cards log, etc.) that describe the mobility and of individuals in the city. Available in large quantities, sampled precisely, but containing few semantics and a lot of noise, they allow a monitoring of the individuals's mobility in the medium term. During this thesis, we propose to work on the modeling of users and the network on the one hand, and the detection of anomalies on the other hand. We will do so using data collected automatically in a context of urban transport networks and using machine learning methods. Moreover, we will focus on the design of methods suited to deal with the particularities of mobility data. We will see that the user-oriented modeling of a transport network allows to obtain fine and robust profiles that can be aggregated efficiently in order to obtain a more precise and more descriptive valuation of the network than a network-oriented modeling. Then, we will explain that the use of these profiles makes it possible to handle complex tasks such as anomaly detection or partitioning of network stations. Finally we will show that the contextualization of the models (spatial context, temporal, shared behaviors) improves the quantitative and qualitative performances
Frédéric, Schmidt. "Classification de la surface de Mars par imagerie hyperspectrale OMEGA. Suivi spatio-temporel et étude des dépôts saisonniers de CO2 et H2O". Phd thesis, 2007. http://tel.archives-ouvertes.fr/tel-00192298.
Les régions polaires de Mars sont le siège d'un cycle climatique annuel d'échange de CO2 entre atmosphère et surface. Pendant la nuit polaire, le CO2 atmosphérique se condense au sol, tandis qu'il se sublime à nouveau pour gonfler l'atmosphère, dès les premiers rayons du soleil au printemps. Ce cycle a été mis à jour depuis les années 60 mais aujourd'hui encore, le détail microphysique d'interaction entre atmosphère et surface demeure inconnu. Le second objectif de cette thèse est d'établir un modèle de sublimation des dépôts saisonniers martiens. Le bilan de masse est simulé par un bilan radiatif sur une surface rugueuse. La confrontation de ce modèle avec différents jeux de données spatiales a permis de montrer que la sublimation de la calotte saisonnière sud de Mars est contrôlée majoritairement par son albédo. Des études ultérieures seront nécessaires pour saisir quels sont les mécanismes à l'origine des variabilités d'albédo (métamorphisme, contamination en poussière, . . . ).