15771 research outputs found
Sort by
Evaluation d’un système d’information et de connaissance - De l’importance de la prise en compte de la connaissance
No full textUn système d’information (SI) est un ensemble organisé de ressources incluant nonseulement des ressources numériques (logiciels, machines, données) mais aussi des ressourceshumaines. Dans notre vision l’humain n’est pas un « simple utilisateur », il est aussi un compo-sant du système, porteur de connaissances. Cette vision nous conduit à introduire le concept deSystème d’Information et de Connaissances (SICO), et à envisager son évaluation. Pour cela,nous choisissons un cadre d’évaluation du système d’information de la littérature, reposant surune vision du SI proche de celle du SICO, que nous étendons afin d’inclure un nouveau pointde vue de l’évaluation : celui de la prise en compte des connaissances. Nous proposons ensuiteun modèle pour l’évaluation de ce point de vue. Ce modèle permet d’apprécier le niveau dematurité de la prise en compte des connaissances au sein d’un SI.ouinonouirechercheNationa
Towards an analytics and an ethics of expertise: Learning from decision aiding experiences in public risk assessment and risk management
No full textPublic expertise in safety, security and environment (SSE) is a process that is increasingly submitted to control and transparency. As decision-making, the exercise of expertise involves subjectivity and judgment. An oversight, a monitoring and an aiding approach is therefore required for its conduct and its governance. This paper proposes a novel way of embedding ethical aspects and participative decision-making elements into the process of risk assessment and risks management. Based on their experience feedback, the authors first propose some early contributions to study the validity and the legitimacy of expertise in SSE. In the second part of the paper, the authors give an overview on how public expertise is organized in France and how the problems and the conclusions are framed in SSE. Finally, the authors propose a generic integrated framework for public expertise that constitutes the “responsible exercise of expertise”. This framework allows framing a valid and a legitimate expertise process and its conclusions.nonnonouirechercheInternationa
Incentive engineering for Boolean games
No full textBoolean games are a natural, compact, and expressive class of logic-based games, in which each player exercises unique control over some set of Boolean variables, and has some logical goal formula that it desires to be achieved. A playerʼs strategy set is the set of all possible valuations that may be made to its variables. A playerʼs goal formula may contain variables controlled by other agents, and in this case, it must reason strategically about how best to assign values to its variables. In the present paper, we consider the possibility of overlaying Boolean games with taxation schemes. A taxation scheme imposes a cost on every possible assignment an agent can make. By designing a taxation scheme appropriately, it is possible to perturb the preferences of agents so that they are rationally incentivised to choose some desirable equilibrium that might not otherwise be chosen, or incentivised to rule out some undesirable equilibria. After formally presenting the model, we explore some issues surrounding it (e.g., the complexity of finding a taxation scheme that implements some desirable outcome), and then discuss possible desirable properties of taxation schemes.nonouirechercheInternationa
Dominance based monte carlo algorithm for preference learning in the multi-criteria sorting problem: Theoretical properties
No full textIn this article we study a new model-free Multi-CriteriaDecision Aiding (MCDA) method for sorting problems where ob-jects are assigned to predefined and ordered categories. The problemthat we deal is the following: given a learning set of objects definedon multi-attributes and already assigned by the decision maker, howto find the assignments of the remaining objects. Being model-free,we do not assume that the decision maker’s reasoning follows somewell-known and explicitly described rules or logic system. We onlyassume that monotonicity should be respected as well as the learningset. The specificity of our approach is to be probabilistic. A MonteCarlo principle is used where the median operator aggregates the re-sults of independent and randomized experiments. We proved thatour final sortings respect the monotonicity and the learning set andthe aggregation with the median operator converges almost surely.nonouirechercheInternationa
Computation of Gaussian orthant probabilities in high dimension
No full textWe study the computation of Gaussian orthant probabilities, i.e. the probability that a Gaussian variable falls inside a quadrant. The Geweke–Hajivassiliou–Keane (GHK) algorithm (Geweke, Comput Sci Stat 23:571–578 1991, Keane, Simulation estimation for panel data models with limited dependent variables, 1993, Hajivassiliou, J Econom 72:85–134, 1996, Genz, J Comput Graph Stat 1:141–149, 1992) is currently used for integrals of dimension greater than 10. In this paper, we show that for Markovian covariances GHK can be interpreted as the estimator of the normalizing constant of a state-space model using sequential importance sampling. We show for an AR(1) the variance of the GHK, properly normalized, diverges exponentially fast with the dimension. As an improvement we propose using a particle filter. We then generalize this idea to arbitrary covariance matrices using Sequential Monte Carlo with properly tailored MCMC moves. We show empirically that this can lead to drastic improvements on currently used algorithms. We also extend the framework to orthants of mixture of Gaussians (Student, Cauchy, etc.), and to the simulation of truncated Gaussians.nonouirechercheInternationa
On the Approximability of Partial VC Dimension
No full textin Springer series Lecture Notes in Computer Science, vol. 10043We introduce the problem Partial VC Dimension that asks, given a hypergraph H=(X,E) and integers k and ℓ, whether one can select a set C⊆X of k vertices of H such that the set {e∩C,e∈E} of distinct hyperedge-intersections with C has size at least ℓ. The sets e∩C define equivalence classes over E. Partial VC Dimension is a generalization of VC Dimension, which corresponds to the case ℓ=2k, and of Distinguishing Transversal, which corresponds to the case ℓ=|E| (the latter is also known as Test Cover in the dual hypergraph). We also introduce the associated fixed-cardinality maximization problem Max Partial VC Dimension that aims at maximizing the number of equivalence classes induced by a solution set of k vertices. We study the approximation complexity of Max Partial VC Dimension on general hypergraphs and on more restricted instances, in particular, neighborhood hypergraphs of graphs.nonouirechercheInternationa
Sequential Halving Applied to Trees
No full textMonte Carlo tree search (MCTS) is state of the art for multiple games and problems. The base algorithm currently used for MCTS is UCT. We propose an alternative MCTS algorithm: sequential halving applied to Trees (SHOT). It has multiple advantages over UCT: it spends less time in the tree, it uses less memory, it is parameter free, at equal time settings it beats UCT for a complex combinatorial game and it can be efficiently parallelized.nonouirechercheInternationa
Efficient Semantic Tableau Generation for Abduction in Propositional Logic
No full textin Frontiers in Artificial Intelligence and Applications, vol. 285Abduction is a backward chaining inference, finding the best explanations of an observation with regard to a knowledge base in a two-steps process: i) hypotheses generation, and ii) explanations selection according to a minimality criterion.nonouirechercheInternationa
Entre l’environnementalisme, le libéralisme et le socialisme : Une recherche sur les pratiques et les idéologies dans la comptabilité environnementale agricole de la Chine
No full textÀ la lumière des perspectives critiques sur les théories et les pratiques comptables, « la comptabilité pour la durabilité » a émergé comme une discipline relativement nouvelle. Par rapport aux différents modèles développés dans le cadre de l’idéologie néolibérale, certains modèles alternatifs avec des stratégies « plus fortes » de conservation des capitaux, notamment la « Triple Ligne d’Amortissement » (TLA) (Richard, 2012; Rambaud et Richard, 2013), semblent avoir beaucoup des potentiels de faire quelques changements fondamentaux. Cette thèse est une enquête sur les conditions nécessaires et les résistances possibles rencontrées par la mise en place du modèle TLA. Le secteur agricole de la République populaire de Chine est choisi comme le terrain de recherche en vue d’étudier le contexte institutionnel et historique. L’enquête sur le contexte chinois est effectuée en référence aux théories de l’économie politique institutionnaliste, inspiré par Karl Polanyi (2002[1944]), Mancur Olson (1965), and Chang Ha-Joon (1994, 2002).In the light of critical perspectives on accounting theories and practices, sustainability accounting (SA) as a relatively new discipline has emerged. Compared to various SA models developed under the mainstream neoliberal ideology, some alternative models with “stronger” capital conservation strategies, notably the “Triple Deprecation Line” (TDL) (Richard, 2012; Rambaud & Richard, 2013), seem to have a lot of potentials to make some fundamental changes. This dissertation is essentially an investigation of necessary conditions and possible resistances faced by the establishment of the TDL model. The agriculture sector of the People’s Republic of China is chosen as the field of research in order to study the institutional environment and historical context in real-world settings. The investigation on the Chinese context is conducted with reference to theories of institutionalist political economy, inspired by Karl Polanyi (2002[1944]), Mancur Olson (1965), and Chang Ha-Joon (1994, 2002)