A5077171510- Data Mining Algorithms and Applications
- Data Management and Algorithms
- Rough Sets and Fuzzy Logic
- Advanced Database Systems and Queries
- Time Series Analysis and Forecasting
- Geographic Information Systems Studies
- Remote Sensing and Land Use
- Semantic Web and Ontologies
- Natural Language Processing Techniques
- Intelligent Tutoring Systems and Adaptive Learning
- French Urban and Social Studies
- Constraint Satisfaction and Optimization
- Species Distribution and Climate Change
- Data Visualization and Analytics
- Plant biochemistry and biosynthesis
- Geochemistry and Geologic Mapping
- Advanced Computational Techniques and Applications
- Multi-Agent Systems and Negotiation
- Online Learning and Analytics
- Microbial Metabolic Engineering and Bioproduction
- Algorithms and Data Compression
- Imbalanced Data Classification Techniques
- Water Systems and Optimization
- Peer-to-Peer Network Technologies
- Environmental DNA in Biodiversity Studies
Centre National de la Recherche Scientifique
2006-2025
Aix-Marseille Université
2023-2025
Laboratoire d’Informatique et Systèmes
2023-2025
University of New Caledonia
2012-2021
Laboratoire d'Informatique en Images et Systèmes d'Information
2008-2010
Université Claude Bernard Lyon 1
2008-2009
Institut National des Sciences Appliquées de Lyon
2009
Clermont Université
2006
Laboratoire d'Informatique, de Modélisation et d'Optimisation des Systèmes
2004-2006
The discovery of frequent patterns is a famous problem in data mining. While plenty algorithms have been proposed during the last decade, only few contributions tried to understand influence datasets on behavior. Being able explain why certain are likely perform very well or poorly some still an open question. In this setting, we describe thorough experimental study with respect item sets. We distribution sets size together three concise representations: closed, free and essential For each...
Extraction of interesting colocations in geo-referenced data is one the major tasks spatial pattern mining. The goal to find sets object-types with instances located same neighborhood. In this context, main drawback visualization and interpretation extracted patterns by domain experts. Indeed, common textual representation loses important information such as position, orientation or distribution patterns. To overcome problem, we propose a new clustering-based technique deeply integrated...
Data mining methods extract knowledge from huge amounts of data. Recently with the explosion mobile technologies, a new type data appeared. The resulting databases can be described as spatiotemporal in which spatial information (e.g., location an event) and temporal (e .g., date are included. In this article, we focus on patterns extraction kind databases. These considered sequences representing changes events localized areas its near surrounding over time. Two algorithms proposed to tackle...
The protection and the maintenance of exceptional environment New Caledonia are major goals for this territory. Among environmental problems, erosion has a strong impact on terrestrial coastal ecosystems. However, due to volume data its complexity, assessment hazard at regional scale is time-consuming, costly rarely updated. Therefore, understanding predicting phenomenons need advanced techniques analysis modelization. In order improve phenomenon, paper proposes spatial approach based...
Spatial data mining has been extensively studied for GIS applications. To deal with a fast increasing of data, investigations spatial analysis are needed. In this paper, we propose approach which adapts the existing colocation concept to characterize soil erosion hazard. order manage task, put task into more general framework. Based on framework, new constraints linked domain knowledge pushed algorithm. Finally, developed prototype and lead experiments real scientific datasets.