A5050957296- Bayesian Modeling and Causal Inference
- Data Management and Algorithms
- Rough Sets and Fuzzy Logic
- Assistive Technology in Communication and Mobility
- Multi-Criteria Decision Making
- Context-Aware Activity Recognition Systems
- Artificial Intelligence in Healthcare
- Fault Detection and Control Systems
- Stroke Rehabilitation and Recovery
- Machine Learning and Data Classification
- semigroups and automata theory
- AI-based Problem Solving and Planning
- Parkinson's Disease Mechanisms and Treatments
- Semantic Web and Ontologies
- Data Quality and Management
- Autism Spectrum Disorder Research
- Safety Warnings and Signage
- Linguistics and Discourse Analysis
- Fire Detection and Safety Systems
- Scheduling and Optimization Algorithms
- Metaheuristic Optimization Algorithms Research
- Health, Medicine and Society
- Digital Accessibility for Disabilities
- Advanced Data Processing Techniques
- Occupational Health and Safety Research
Laboratoire d'Automatique, de Mécanique et d'Informatique Industrielles et Humaines
2015-2024
Université Polytechnique Hauts-de-France
2006-2023
Centre National de la Recherche Scientifique
2009-2023
Université Lille Nord de France
2009-2014
Université de Lille
2009-2011
This paper aims to get a better understanding of the notions evidence, probabilistic evidence and likelihood in Bayesian Networks. Evidence comes from an observation one or several variables. Soft is since consists local probability distribution on subset variables that has replace any former belief these It be clearly distinguished also called virtual for which specified as ratio. Since notion soft not yet widely understood, most Networks engines do propose related propagation functions...
Multiple diagnosis methods using Bayesian networks are rooted in numerous research projects about model-based diagnosis. Some of this exploits probabilities to make a Many network applications used for medical or the technical problems small moderately large devices. This paper explains detail advantages as graphic probabilistic models diagnosing complex devices, and then compares such with other that may not use networks.
The Bayesian Networks are graphical models that easy to interpret and update. These useful if the knowledge is uncertain, but they lack some means express ambiguity. To face this problem, we propose Fuzzy Evidence in combine Logic Network. This has allowed benefit from mutual advantages of these two approaches, overcome problem data observation paper proposes an inference algorithm which uses Network reliability. solution been implemented, tested evaluated comparison with existing methods.
The multi-criteria decision problem is to find the most satisfactory solution among many alternatives, taking into account several criteria that may be conflicting. proposed method focuses on person since importance (weight) given each criterion defined according characteristics of person. We use a special structure Bayesian network (BN) based AHP. graph BN and probabilities associated with nodes are designed translate knowledge experts selection an alternative.
Falls are frequent in the elderly and it is number one cause of traumatic death this population. Fall prevention requires to evaluate which risk factors for fall present a person on basis incomplete health data. In context, main objective paper identify falls people them based partial observations. Health data study was provided by hospital Lille, France. Firstly, identified from with help an ontology fall. Furthermore, steps pre-processing, missing value imputation variable selection...
There are several situations where humans express their preferences in order to take good decisions. The major problem is that humans' more and complex, the multiple criteria considered often conflicting number of alternatives too large be explicitly handled. objective Multi-Criteria Decision Making (MCDM) approaches efficiently model solve such complex decision problems. In this paper, we propose a framework allowing on one hand encode users' about regarding available using logic-based...
In this paper, we focus on multi-criteria decision-making problems. We propose a model based influence diagrams; is able to handle uncertainty, represent interdependencies among the different decision variables and facilitate communication between decision-maker analyst. The particular structure of proposed makes it possible take into account alternatives described by an attribute set, decision-maker's characteristics preferences, other information (e.g., internal or external factors) that...
Cortical activity and walking speed are known to decline with age can lead an increased risk of falls in the elderly. Despite being a contributor this decline, individuals at different rates. This study aimed analyse left right cortical changes elderly adults regarding their speed. activation gait data were obtained from 50 healthy older individuals. Participants then grouped into cluster based on preferred (slow or fast). Analyses differences parameters between groups carried out....
Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI
Background: Falls in the elderly are number one cause of traumatic death this population. Prevention falls requires to evaluate which risk factors for fall present a person on basis available health information. Our objective is predict presence or absence 12 people based partial observations. Methods: A data set 1810 patients multidisciplinary consultation Lille University Hospital covering fourteen years admissions were used learn and Bayesian network four usual machine learning...