Dhouha Grissa
PhD. in computer science
Research
Principal research work
My principal researches have focused on the development of a strategy based on the use of knowledge extraction methods and metabolomic data mining (global analysis of small molecules in a biological system) to facilitate the discovery of biomarkers of change in health conditions within a cohort study. It is ambitious and innovative to analyze, explore, visualize and interpret large amounts of complex biological data, as part of a major public health problem, obesity and its complications to diabetes type 2. Identification of individuals or populations at risk of developing metabolic diseases and improve their health is the major goal of my current research. In this project, the volume of data is important and data are as well heterogeneous (both numerical and symbolic). The integration of large volumes of data can be guided by domain knowledge and be supported by a data schema considered as a mediation system (virtual integration needing correspondences between data sources). In this context, my research consists in studying the set of data to be analyzed from a theoretical and practical point of view.
During my thesis, I worked on the behavioral of interestingness measures of knowledge extraction, especially on objective measures (data-driven) theoretical and practical point of view. The purpose of my work was to help the user making the right choice of the most appropriate interestingness measure to his application. Unsupervised clustering methods have been used in order identify groups of interestingness measures with similar behavior.
Laboratories affiliation: NNF Center for Protein Research - Jensen group-
Projects & collaborations
Research projects
-
Galaxy (Gut-and-liver axis in alcoholic liver fibrosis). European Project, Grant Number: 668031, responsible Lars Juhl Jensen (University of Copenhagen).
-
DIAPASON (Diet-health InterAction throughout life - Predictive biomArkers of life tranSitiON outcome related to retirement). Metaprogram INRA, Did’it, responsible Blandine Comte.
-
PEPS HyQual (A Hybrid and Qualitative approach for mining complex biological data). CNRS PEPS, responsible Amedeo Napoli.
-
PHC Utique EXQUI No. 11G1417: Franco-Tunisian cooperation. EXQUI (EXtraction, QUality and engineering knowledge in heterogeneous environments) http://home.isima.fr/exqui/
​
Summer schools and workshops participation
-
Participation to INRA BioInfo journey 2016: INRA Bioinformatics journeys, 21-23 Mar. 2016, Toulouse France (Poster presentation).
-
Participation to the french journey JCPU 2014: Clermontoise journeys of teaching university, 3-4 juil. 2014, Clermont-Ferrand French (Poster presentation).
-
Participation to the international summer school SSIU 2012: Information and Uncertainty, 4 - 8 june 2012, Olomouc, Czk republic.
-
Participation to the international workshop WIUI 2012: Information, Uncertainty, and Imprecision, 5-7 june 2012, Olomouc, Czk republic (Poster presentation).​
Publications
​
-
Links:
Google Scholar: https://scholar.google.com/citations?user=7q2VeQ4AAAAJ&hl=fr
Researchgate: https://www.researchgate.net/profile/Dhouha_Grissa
​
-
PhD Thesis:
Dhouha GRISSA (defense 02/12/2013). "Étude comportementale de mesures d’intérêt d'extraction de connaissances". Phd thesis, University of Blaise-Pascal, France and University of Tunis-El Manar, Tunisia.
-
In international journals with referees:
Dhouha GRISSA, Mélanie PETERA, Marion BRANDOLINI, Amedeo NAPOLI, Blandine COMTE and Estelle PUJOS-GUILLOT (2016). "Feature selection methods for early predictive biomarker discovery using untargeted metabolomic data". In Journal FRONTIERS in Molecular Biosciences - Metabolomics.
Radim BELOHLAVEK, Dhouha GRISSA, Sylvie GUILLAUME, Engelbert MEPHU NGUIFO and Jan OUTRATA (2013). "Boolean factors as a means of clustering of interestingness measures of association rules". In AMAI Journal, Springer Netherlands.
-
In national journals with referees:
Sylvie GUILLAUME, Dhouha GRISSA and Engelbert MEPHU NGUIFO (2011). "Catégorisation des mesures d'intérêt pour l'extraction des connaissances". In Revue des Nouvelles Technologies de l'Information. p. 117-144.
-
In international conferences (also French-speaking) with referees:
Dhouha GRISSA, Blandine COMTE, Estelle PUJOS-GUILLOT and Amedeo NAPOLI (2016). "A Hybrid Knowledge Discovery Approach for Mining Predictive Biomarkers in Metabolomic data". In European Conf. on Machine Learning and Principles and Practice of Knowledge Discovery (ECML/PKDD’16 (1)). Italy, 572-587.
Dhouha GRISSA, Blandine COMTE, Estelle PUJOS-GUILLOT and Amedeo NAPOLI (2016). "A Hybrid Approach for Mining Metabolomic data". In 5th workshop "What can FCA do for Artificial Intelligence" (FCA4AI/ECAI’16). The Hague, the Netherlands.
Dhouha GRISSA, Blandine COMTE, Estelle PUJOS-GUILLOT and Amedeo NAPOLI (2016). "A Hybrid Data Mining Approach for the Identification of Biomarkers in Metabolomic data". In Int. Conf. Concept Lattice and their Applications (CLA’2016). Moscow, Russia, 161-174.
Dhouha GRISSA, Blandine COMTE, Estelle PUJOS-GUILLOT and Amedeo NAPOLI (2016). "Combinaison de méthodes symboliques et numériques pour l'analyse de données métabolomiques". In Proceed. Int. Conf. Extraction et Gestion des Connaissances (EGC’2016). Reims, France, 555-556.
Radim BELOHLAVEK, Dhouha GRISSA, Sylvie GUILLAUME, Engelbert MEPHU NGUIFO and Jan OUTRATA (2011). "Boolean factors as a means of clustering of interestingness measures of association rules". In CLA’2011, Nancy-FRANCE. p.207-222.
Sylvie GUILLAUME, Dhouha GRISSA et Engelbert MEPHU NGUIFO (2011). "Catégorisation des mesures d’intérêt pour l’extraction des connaissances". In EGC’2011, p. 551-562.
-
In national conferences with referees:
Dhouha GRISSA, Jérémie BOURSEAU, Blandine COMTE, Estelle PUJOS-GUILLOT and Amedeo NAPOLI (2015). "Knowledge Discovery based on Formal Concept Analysis for biomarker identification from metabolomics data". In Journées Ouvertes en Biologie, Informatique et Mathématiques (JOBIM’2015). France. 2015. (Poster)
Dhouha GRISSA (2013). "Étude comportementale de mesures d’intérêt de règles d’association". Dans RJCIA, PFIA’2013, Lille-France.
Dhouha GRISSA, Sylvie GUILLAUME, Engelbert MEPHU NGUIFO et Sadok BEN YAHIA (2011). "Étude expérimentale des mesures d'intérêt pour l'extraction des connaissances". In Extraction et Contextualisaton des Connaissances workshop, AFIA’2011, Chambéry-FRANCE.
Sylvie GUILLAUME, Dhouha GRISSA et Engelbert MEPHU NGUIFO (2010). "Propriétés des mesures d’intérêt pour l’extraction des règles". In Qualité des Données et des Connaissances workshop, EGC’2010, Hammamet-Tunisie.
Research fields
data mining and machine learning:
- interestingness measures
- clustering
- supervised classification
- association rules
- boolean factor analysis
- formal properties of measures
- formal concept analysis
Bioinformatics:
- metabolomics data analysis
- feature selection
- biomarker identification
- prediction
- ROC curves
statistic:
- data analysis
- unsupervised clustering
methods