Academia.eduAcademia.edu

Inference in possibilistic networks

description11 papers
group2 followers
lightbulbAbout this topic
Inference in possibilistic networks refers to the process of deriving conclusions or making predictions based on uncertain information represented in a graphical model, where nodes represent variables and edges denote relationships, utilizing possibility theory to handle imprecise or vague data.
lightbulbAbout this topic
Inference in possibilistic networks refers to the process of deriving conclusions or making predictions based on uncertain information represented in a graphical model, where nodes represent variables and edges denote relationships, utilizing possibility theory to handle imprecise or vague data.

Key research themes

1. How can new algorithmic frameworks improve exact and approximate inference efficiency in discrete probabilistic graphical models?

This research area focuses on developing novel algorithms that enable efficient exact and approximate inference in probabilistic graphical models—especially discrete Bayesian networks and Markov chains—by exploiting structural properties and computational mechanisms such as generators and join-tree propagation. The motivation is to overcome computational bottlenecks in conventional inference methods to better handle complex, large-scale discrete domains.

Key finding: Introduces the Statues algorithm, a novel exact inference method for discrete probabilistic models expressed as directed acyclic graphs, which leverages generator constructs (coroutines) to significantly reduce unnecessary... Read more
Key finding: Establishes that the Simple Propagation (SP) algorithm, a join-tree method for exact inference in discrete Bayesian networks, accurately identifies relevant potentials equivalently to established Lazy Propagation (LP)... Read more
Key finding: Proposes the use of value-based potentials (VBPs), data structures that exploit repetition of values in probabilistic potentials regardless of their ordering, to approximate and efficiently represent conditional probability... Read more

2. How do independence criteria and conditional independence tests affect the accuracy of network structure inference from observational data?

This theme examines the role of independence measures—ranging from classic linear-Gaussian tests to more generalized statistics like Hilbert-Schmidt Independence Criterion (HSIC) and Distance Covariance Criterion (DCC)—in inferring network structures, notably directed acyclic graphs, from data. Central to this theme is improving causal or functional dependency identification by exploiting nonlinear dependencies and non-Gaussian noise, addressing limitations in traditional methods that assume linearity and Gaussianity.

Key finding: Demonstrates that generalized independence criteria such as HSIC and Distance Covariance lead to improved inference of directed acyclic graph structures over standard linear-Gaussian tests when used in the PC algorithm... Read more
Key finding: Surveys methods for learning causal and belief networks emphasizing the particular challenges of recovering causal structures under possibilistic frameworks. Presents new results on algorithms that extend belief network... Read more
Key finding: Critically reviews the classical belief propagation algorithm and its loopy variant (LBP) for inference on Bayesian networks with cycles, indicating that while LBP often provides good approximations in practice, it may... Read more

3. What are the challenges and approaches in learning and validating Bayesian network structures from domain-specific expert knowledge, arguments, or passively collected data?

This research area investigates methods to derive Bayesian network structures either from structured arguments or passive data collection, including adversary perspective modeling and network reconstruction with feasibility guarantees. Challenges include bridging expert knowledge represented as arguments to formal probabilistic models, ensuring inferred networks accurately explain input data or traces, and constructing interpretable decision support systems that model belief formation.

Key finding: Proposes a systematic method to extract structural and probabilistic constraints on Bayesian network models from structured arguments, enabling checking compatibility of existing networks with domain expert reasoning or... Read more
Key finding: Develops a novel Bayesian inference method for social network reconstruction which incorporates constraints ensuring that the inferred network is trace-feasible—that is, every observed interaction in the data can be causally... Read more
Key finding: Presents a model-driven decision support system utilizing Bayesian belief networks to simulate adversarial inference techniques from passively collected network data, demonstrated with operating system fingerprinting. The... Read more

All papers in Inference in possibilistic networks

We describe an Information Retrieval Model based on fuzzy Networks that incorporates dependence relationships between indexing terms. We details the design and implementation of a new Information Retrieval Model based on Fuzzy Network.... more
We propose a new approach denoted BNDI (Bayesian Network for Document Indexing) for indexing biomedical documents with controlled biomedical vocabulary based on a Bayesian Network. BNDI uses the probability inference to extract... more
This paper focuses on probability updates in multiply-connected belief networks. Pearl has designed the method of conditioning, which enables us to apply his algorithm for belief updates in singly-connected networks to multiply-connected... more
The First International Conference on Information Processing and the Management of Uncertainty (IPMU) was held in 1986 at a time of great debate about the necessity of modelling uncertainty in intelligent systems (which at that time... more
Dans ce papier, nous présentons les méthodes que nous avons développées pour participer aux tâches 1 et 2 de l’édition 2019 du défi fouille de textes (DEFT 2019). Pour la première tâche, qui s’intéresse à l’indexation de cas cliniques,... more
Background: The situation of medical coding and medical economics is quite specific in France. Two specific health terminologies are used : ICD10 and CCAM. Another will be used in the near future SNOMED Int. Objective: The objective of... more
This paper presents an Internet information retrieval system based on Hierarchical Small-Worlds (HSW) and Possibilistic Networks (PN). The first HSW, for the words of the French language, is used to take account of the dependences between... more
Le cerveau humain est capable d'identifier les thématiques d'un texte en le parcourant, sans vraiment chercher à comprendre son contenu. C'est ce que cherche à faire la procédure décrite dans cet article. Elle procède en deux temps. Tout... more
Le cerveau humain est capable d'identifier les thématiques d'un texte en le parcourant, sans vraiment chercher à comprendre son contenu. C'est ce que cherche à faire la procédure décrite dans cet article. Elle procède en deux temps. Tout... more
RESUME. Nous proposons dans cet article une nouvelle approche d'indexation de documents biomédicaux basée sur les réseaux possibilistes permettant de les apparier partiellement aux termes du thésaurus MeSH (Medical Subject Headings). La... more
Dans cet article, nous présentons nos méthodes pour les tâches d’indexation et d’appariements du Défi Fouile de Textes (Deft) 2019. Pour la taĉhe d’indexation nous avons testé deux méthodes, une fondée sur l’appariemetn préalable des... more
The Supervisory Control and Data Acquisition (SCADA) system is the most commonly used industrial control system but is subject to a wide range of serious threats. Intrusion detection systems are deployed to promote the security of SCADA... more
RESUME. Nous proposons dans cet article une nouvelle approche d'indexation de documents biomédicaux basée sur les réseaux possibilistes permettant de les apparier partiellement aux termes du thésaurus MeSH (Medical Subject Headings). La... more
Dans le domaine medical, l'informatisation des professions de sante et le developpement du dossier medical personnel (DMP) entraine une progression rapide du volume d'information medicale numerique. Le besoin de convertir et de... more
Background: The situation of medical coding and medical economics is quite specific in France. Two specific health terminologies are used : ICD10 and CCAM. Another will be used in the near future SNOMED Int. Objective: The objective of... more
Dans ce papier, nous présentons les méthodes que nous avons développées pour participer aux tâches 1 et 2 de l’édition 2019 du défi fouille de textes (DEFT 2019). Pour la première tâche, qui s’intéresse à l’indexation de cas cliniques,... more
Dans ce papier, nous présentons les méthodes que nous avons développées pour participer aux tâches 1 et 2 de l’édition 2019 du défi fouille de textes (DEFT 2019). Pour la première tâche, qui s’intéresse à l’indexation de cas cliniques,... more
French pharmaceutical thesis are seldom refered to. If the main obstacles originate from language or access barriers, proper indexation could also be blaimed. Manually extracted key words dont necessary come from a structured thesaurus.... more
Catalogue and Index of French Medical Sites (CISMeF) is developed for retrieving the relevant medical information in the Internet for health professionals, the patients and students in medicine. The gathered resources are manually... more
This thesis deals with the problemn of Information Retrieval on the Internet. We propose several methods of query expansion founded on knowledge exploitation. Experimentations are done in the context of the CISMeF project which indexes... more
peut améliorer son efficacité, en particulier, dans des domaines spécifiques. La plupart des travaux utilisent les concepts comme une alternative aux mots et transforment le classique sac de mots (BOW) en un sac de concepts (BOC). Cette... more
peut améliorer son efficacité, en particulier, dans des domaines spécifiques. La plupart des travaux utilisent les concepts comme une alternative aux mots et transforment le classique sac de mots (BOW) en un sac de concepts (BOC). Cette... more
Background: The situation of medical coding and medical economics is quite specific in France. Two specific health terminologies are used : ICD10 and CCAM. Another will be used in the near future SNOMED Int. Objective: The objective of... more
Bayesian Networks have been proposed as an alternative to rule-based systems in domains with uncertainty. Applications in monitoring and control can benefit from this form of knowledge representation. Following the work of Chong and... more
Bayesian Networks have been proposed as an alternative to rule-based systems in domains with uncertainty. Applications in monitoring and control can benefit from this form of knowledge representation. Following the work of Chong and... more
Dans ce papier, nous présentons les méthodes que nous avons développées pour participer aux tâches 1 et 2 de l’édition 2019 du défi fouille de textes (DEFT 2019). Pour la première tâche, qui s’intéresse à l’indexation de cas cliniques,... more
This paper presents an Internet information retrieval system based on Hierarchical Small-Worlds (HSW) and Possibilistic Networks (PN). The first HSW, for the words of the French language, is used to take account of the dependences between... more
This paper presents an Internet information retrieval system based on Hierarchical Small-Worlds (HSW) and Possibilistic Networks (PN). The first HSW, for the words of the French language, is used to take account of the dependences between... more
In this paper, we are interested in aggregated search in structured XML documents. We present a model for the structured information retrieval, based on the Bayesian networks theory. Relations query-terms and terms-elements are modelled... more
Dans ce papier, nous présentons les méthodes que nous avons développées pour participer aux tâches 1 et 2 de l’édition 2019 du défi fouille de textes (DEFT 2019). Pour la première tâche, qui s’intéresse à l’indexation de cas cliniques,... more
In this paper we present a synthesis of the work pe rformed on two inference algorithms: the Pearl’s belief propagation (BP) algorithm applied t o Bayesian networks without loops (i.e. polytree) and the Loopy belief propagation (LBP) al... more
Les travaux menes dans le cadre de cette these se situent dans la problematique de recherche- indexation des documents et plus specifiquement dans celle de l’extraction des descripteurs semantiques pour l’indexation. Le but de la... more
Dans cet article nous présentons une approche statistique d'indexation sémantique des documents multilingues. Cette approche est validée par un ensemble d'expérimentations et une comparaison avec une approche linguistique. Nous montrons... more
Les travaux menés dans le cadre de cette thèse se situent dans la problématique de recherche- indexation des documents et plus spécifiquement dans celle de l'extraction des descripteurs sémantiques pour l'indexation. Le but de la... more
Dans cet article nous présentons une approche statistique d'indexation sémantique des documents multilingues. L'approche que nous proposons est composée de trois étapes : extraction des termes, détection des concepts et détection des... more
Dans cet article nous présentons une approche statistique d'indexation sémantique des documents multilingues. Cette approche est validée par un ensemble d'expérimentations et une comparaison avec une approche linguistique. Nous montrons... more
La capture de relations sémantiques entre termes à partir de textes est un moyen privilégié de constituer/alimenter une base de connaissances, ressource indispensable pour l’analyse de textes. Nous proposons et évaluons la combinaison de... more
Les données médicales étant de plus en plus informatisées, le traitement sémantiquement efficace des rapports médicaux est devenu une nécessité. La recherche d’images radiologiques peut être grandement facilitée grâce à l’indexation... more
Most artificial intelligence applications, especially expert systems, have to reason and make decisions based on uncertain data and uncertain models. For this reason, several methods have been proposed for reasoning with different kinds... more
In this paper we present a synthesis of the work performed on two inference algorithms: the Pearl's belief propagation (BP) algorithm applied to Bayesian networks without loops (i.e. polytree) and the Loopy belief propagation (LBP)... more
We present a novel inference algorithm which is an adaptation of Loopy Belief Propagation applied on Product-Based Possibilistic Networks. Without any transformation of the initial graph, the basic idea of this adaptation is to propagate... more
We present a novel inference algorithm which is an adaptation of Loopy Belief Propagation applied on Product-Based Possibilistic Networks. Without any transformation of the initial graph, the basic idea of this adaptation is to propagate... more
Dans cet article, nous présentons un modèle pour la recherche d'information structurée en XML, basé sur les réseaux possibilistes. Les relations document-éléments et éléments-termes sont modélisées par des mesures de possibilité et de... more
Les entrepôts de données contiennent des données sensibles qui doivent être protégées contre les accès non autorisés, aussi bien directs que par inférence. Les accès directs sont contrôlables par des autorisations gérées par le serveur... more
Cet article décrit la banque documentaire MEDLINE depuis laquelle une collection test comprenant environ 4,5 million de documents structurés a été construite à partir des campagnes d'évaluation TREC. Dans une deuxième partie, nous... more
The situation of medical coding and medical economics is quite specific in France. Two specific health terminologies are used : ICD10 and CCAM. Another will be used in the near future SNOMED Int. Objective: The objective of this study is... more
La croissance rapide de la quantité d’information a motivé l’émergence de nombreux domaines infor- matiques, dont les systèmes de résumé automatique (RA). Ces systèmes sont utilisés pour extraire les informations les plus... more
Download research papers for free!