Bien que la recherche ait fait progresser la traduction automatique depuis plusieurs années, la sortie d’un système automatisé ne peut être généralement publiée sans avoir été révisée humainement au préalable, et corrigée le cas échéant. Forts de ce constat, nous avons voulu exploiter ces retours utilisateurs issus du processus de révision pour adapter notre système statistique dans le temps, au moyen d’une approche incrémentale.Dans le cadre de cette thèse Cifre-Défense, nous nous sommes donc intéressés à la postédition, un des champs de recherche les plus actifs du moment, et qui plus est très utilisé dans l’industrie de la traduction et de la localisation.L’intégration de retours utilisateurs n’est toutefois pas une tâche aussi évidente qu’il n’y paraît. D’une part, il faut être capable d’identifier l’information qui sera utile au système, parmi l’ensemble des modifications apportées par l’utilisateur. Pour répondre à cette problématique, nous avons introduit une nouvelle notion (les « Actions de Post-Édition »), et proposé une méthodologie d’analyse permettant l’identification automatique de cette information à partir de données post-éditées. D’autre part, concernant l’intégration continue des retours utilisateurs nous avons développé un algorithme d’adaptation incrémentale pour un système de traduction statistique, lequel obtient des performances supérieures à la procédure standard. Ceci est d’autant plus intéressant que le développement et l’optimisation d’un tel système de traduction estune tâche très coûteuse en ressources computationnelles, nécessitant parfois jusqu’à plusieurs jours de calcul.Conduits conjointement au sein de l’entreprise SYSTRAN et du LIUM, les travaux de recherche de cette thèse s’inscrivent dans le cadre du projet ANR COSMAT 1. En partenariat avec l’INRIA, ce projet avait pour objectif de fournir à la communauté scientifique un service collaboratif de traduction automatique de contenus scientifiques. Outre les problématiques liéesà ce type de contenu (adaptation au domaine, reconnaissance d’entités scientifiques, etc.), c’est l’aspect collaboratif de ce service avec la possibilité donnée aux utilisateurs de réviser les traductions qui donne un cadre applicatif à nos travaux de recherche. / Although machine translation research achieved big progress for several years, the output of an automated system cannot be published without prior revision by human annotators. Based on this fact, we wanted to exploit the user feedbacks from the review process in order to incrementally adapt our statistical system over time.As part of this thesis, we are therefore interested in the post-editing, one of the most active fields of research, and what is more widely used in the translation and localization industry.However, the integration of user feedbacks is not an obvious task. On the one hand, we must be able to identify the information that will be useful for the system, among all changes made by the user. To address this problem, we introduced a new concept (the “Post-Editing Actions”), and proposed an analysis methodology for automatic identification of this information from post-edited data. On the other hand, for the continuous integration of user feedbacks, we havedeveloped an algorithm for incremental adaptation of a statistical machine translation system, which gets higher performance than the standard procedure. This is even more interesting as both development and optimization of this type of translation system has a very computational cost, sometimes requiring several days of computing.Conducted jointly with SYSTRAN and LIUM, the research work of this thesis is part of the French Government Research Agency project COSMAT 2. This project aimed to provide a collaborative machine translation service for scientific content to the scientific community. The collaborative aspect of this service with the possibility for users to review the translations givesan application framework for our research.

A swarm intelligence system is a type of multiagent system with the following distinctive characteristics: (i) it is composed of a large number of agents, (ii) the agents that comprise the system are simple with respect to the complexity of the task the system is required to perform, (iii) its control relies on principles of decentralization and self-organization, and (iv) its constituent agents interact locally with one another and with their environment. <p><p>Interactions among agents, either direct or indirect through the environment in which they act, are fundamental for swarm intelligence to exist; however, there is a class of interactions, referred to as "interference", that actually blocks or hinders the agents' goal-seeking behavior. For example, competition for space may reduce the mobility of robots in a swarm robotics system, or misleading information may spread through the system in a particle swarm optimization algorithm. One of the most visible effects of interference in a swarm intelligence system is the reduction of its efficiency. In other words, interference increases the time required by the system to reach a desired state. Thus, interference is a fundamental problem which negatively affects the viability of the swarm intelligence approach for solving important, practical problems.<p><p>We propose a framework called "incremental social learning" (ISL) as a solution to the aforementioned problem. It consists of two elements: (i) a growing population of agents, and (ii) a social learning mechanism. Initially, a system under the control of ISL consists of a small population of agents. These agents interact with one another and with their environment for some time before new agents are added to the system according to a predefined schedule. When a new agent is about to be added, it learns socially from a subset of the agents that have been part of the system for some time, and that, as a consequence, may have gathered useful information. The implementation of the social learning mechanism is application-dependent, but the goal is to transfer knowledge from a set of experienced agents that are already in the environment to the newly added agent. The process continues until one of the following criteria is met: (i) the maximum number of agents is reached, (ii) the assigned task is finished, or (iii) the system performs as desired. Starting with a small number of agents reduces interference because it reduces the number of interactions within the system, and thus, fast progress toward the desired state may be achieved. By learning socially, newly added agents acquire knowledge about their environment without incurring the costs of acquiring that knowledge individually. As a result, ISL can make a swarm intelligence system reach a desired state more rapidly. <p><p>We have successfully applied ISL to two very different swarm intelligence systems. We applied ISL to particle swarm optimization algorithms. The results of this study demonstrate that ISL substantially improves the performance of these kinds of algorithms. In fact, two of the resulting algorithms are competitive with state-of-the-art algorithms in the field. The second system to which we applied ISL exploits a collective decision-making mechanism based on an opinion formation model. This mechanism is also one of the original contributions presented in this dissertation. A swarm robotics system under the control of the proposed mechanism allows robots to choose from a set of two actions the action that is fastest to execute. In this case, when only a small proportion of the swarm is able to concurrently execute the alternative actions, ISL substantially improves the system's performance. / Doctorat en Sciences de l'ingénieur / info:eu-repo/semantics/nonPublished

This thesis describes a combination of two current areas of research: the Crayons image classifier system and active learning. Currently Crayons provides no guidance to the user in what pixels should be labeled or when the task is complete. This work focuses on two main areas: 1) active learning for user guidance, and 2) accuracy estimation as a measure of completion. First, I provide a study through simulation and user experiments of seven active learning techniques as they relate to Crayons. Three of these techniques were specifically designed for use in Crayons. These three perform comparably to the others and are much less computationally intensive. A new widget is proposed for use in the Crayons environment giving an overview of the system "confusion". Second, I give a comparison of four accuracy estimation techniques relating to true accuracy and for use as a completion estimate. I show how three traditional accuracy estimation techniques are ineffective when placed in the Crayons environment. The fourth technique uses the same computation as the three new active learning techniques proposed in this work and thus requires little extra computation and outstrips the other three as a completion estimate both in simulation and user experiments.

In this work, we propose a specialization of the inductive incremental coverability algorithm that solves alternating finite automata emptiness problem. We experiment with various design decisions, analyze them and prove their correctness. Even though the problem itself is PSpace-complete, we are focusing on making the decision of emptiness computationally feasible for some practical classes of applications. We have obtained interesting comparative results against state-of-the-art algorithms, especially in comparison with antichain-based algorithms.

The aim of the thesis is the design machines for laser cutting and plasma workspace 2.0 x1, 8 meters. The work described technology laser cutting and plasma and the resulting demands on the machine. Longer work includes a search of design solutions, then the actual design, including the necessary calculations and drawings.

This diploma thesis consist of overview for position measuring methods and is mainly focused on design of non-standard inductive and incremental sensor. Specifications of those sensors is based on requirements set by MESING company. Also contains verification and testing measurement of completed sensor.

This diploma thesis implements methods for a measurement of rotational speed. It is implemented on the Compact RIO platform from National Instruments. Corresponding SW is implemented using the graphical programming language G in LabView environment. Developed system uses two different sensors – incremental encoder and tacho sensor for measurements of rotational speed. Thesis further analysis and implements a method for an encoder nonlinearity determination and for its on-line correction. For used methods, effects adding errors into the measurements are evaluated and quantified.

An algorithm of incremental mining implicative logical rules is pro-posed. This algorithm is based on constructing good classification tests. The in-cremental approach to constructing these rules allows revealing the interde-pendence between two fundamental components of human thinking: pattern recognition and knowledge acquisition.

In this thesis, we study the notion of graph summarization, which is a fundamental task of finding a compact representation of the original graph called the summary. Graph summarization can be used for reducing the footprint of the input graph, better visualization, anonymizing the identity of users, and query answering. There are two different frameworks of graph summarization we consider in this thesis, the utility-based framework and the correction set-based framework. In the utility-based framework, the input graph is summarized until a utility threshold is not violated. In the correction set-based framework a set of correction edges is produced along with the summary graph. In this thesis we propose two algorithms for the utility-based framework and one for the correction set-based framework. All these three algorithms are for static graphs (i.e. graphs that do not change over time). Then, we propose two more utility-based algorithms for fully dynamic graphs (i.e. graphs with edge insertions and deletions). Algorithms for graph summarization can be lossless (summarizing the input graph without losing any information) or lossy (losing some information about the input graph in order to summarize it more). Some of our algorithms are lossless and some lossy, but with controlled utility loss. Our first utility-driven graph summarization algorithm, G-SCIS, is based on a clique and independent set decomposition, that produces optimal compression with zero loss of utility. The compression provided is significantly better than state-of-the-art in lossless graph summarization, while the runtime is two orders of magnitude lower. Our second algorithm is T-BUDS, a highly scalable, utility-driven algorithm for fully controlled lossy summarization. It achieves high scalability by combining memory reduction using Maximum Spanning Tree with a novel binary search procedure. T-BUDS outperforms state-of-the-art drastically in terms of the quality of summarization and is about two orders of magnitude better in terms of speed. In contrast to the competition, we are able to handle web-scale graphs in a single machine without performance impediment as the utility threshold (and size of summary) decreases. Also, we show that our graph summaries can be used as-is to answer several important classes of queries, such as triangle enumeration, Pagerank and shortest paths. We then propose algorithm LDME, a correction set-based graph summarization algorithm that produces compact output representations in a fast and scalable manner. To achieve this, we introduce (1) weighted locality sensitive hashing to drastically reduce the number of comparisons required to find good node merges, (2) an efficient way to compute the best quality merges that produces more compact outputs, and (3) a new sort-based encoding algorithm that is faster and more robust. More interestingly, our algorithm provides performance tuning settings to allow the option of trading compression for running time. On high compression settings, LDME achieves compression equal to or better than the state of the art with up to 53x speedup in running time. On high speed settings, LDME achieves up to two orders of magnitude speedup with only slightly lower compression. We also present two lossless summarization algorithms, Optimal and Scalable, for summarizing fully dynamic graphs. More concretely, we follow the framework of G-SCIS, which produces summaries that can be used as-is in several graph analytics tasks. Different from G-SCIS, which is a batch algorithm, Optimal and Scalable are fully dynamic and can respond rapidly to each change in the graph. Not only are Optimal and Scalable able to outperform G-SCIS and other batch algorithms by several orders of magnitude, but they also significantly outperform MoSSo, the state-of-the-art in lossless dynamic graph summarization. While Optimal produces always the most optimal summary, Scalable is able to trade the amount of node reduction for extra scalability. For reasonable values of the parameter $K$, Scalable is able to outperform Optimal by an order of magnitude in speed, while keeping the rate of node reduction close to that of Optimal. An interesting fact that we observed experimentally is that even if we were to run a batch algorithm, such as G-SCIS, once for every big batch of changes, still they would be much slower than Scalable. For instance, if 1 million changes occur in a graph, Scalable is two orders of magnitude faster than running G-SCIS just once at the end of the 1 million-edge sequence. / Graduate

Designing object-oriented software is hard, and designing reusable object-oriented software is even harder. This task is even more daunting for a developer of computational intelligence applications, as optimising one design objective tends to make others inefficient or even impossible. Classic examples in computer science include ‘storage vs. time’ and ‘simplicity vs. flexibility.’ Neural network requirements are by their very nature very tightly coupled – a required design change in one area of an existing application tends to have severe effects in other areas, making the change impossible or inefficient. Often this situation leads to a major redesign of the system and in many cases a completely rewritten application. Many commercial and open-source packages do exist, but these cannot always be extended to support input from other fields of computational intelligence due to proprietary reasons or failing to fully take all design requirements into consideration. Design patterns make a science out of writing software that is modular, extensible and efficient as well as easy to read and understand. The essence of a design pattern is to avoid repeatedly solving the same design problem from scratch by reusing a solution that solves the core problem. This pattern or template for the solution has well understood prerequisites, structure, properties, behaviour and consequences. CILib is a framework that allows developers to develop new computational intelligence applications quickly and efficiently. Flexibility, reusability and clear separation between components are maximised through the use of design patterns. Reliability is also ensured as the framework is open source and thus has many people that collaborate to ensure that the framework is well designed and error free. This dissertation discusses the design and implementation of a generic neural network framework that allows users to design, implement and use any possible neural network models and algorithms in such a way that they can reuse and be reused by any other computational intelligence algorithm in the rest of the framework, or any external applications. This is achieved by using object-oriented design patterns in the design of the framework. / Dissertation (MSc)--University of Pretoria, 2007. / Computer Science / unrestricted