On Hierarchical Classification Implicative and Cohesive MGK-Based: Application on Analysis of the Computing Curricula and Students Abilities According the Anglo-Saxon Model

Hery Frédéric Rakotomalala (1) , Totohasina André (1)

1 Ecole Normale Supérieure pour l'enseignement Technique [Antsiranana]

Extracting association rules from a huge binary data according to a quality measure is an important pretreatment step in data analysis.Also, among unsupervised techniques, our approach for a hierarchical classification implicative and cohesive is based on the new measure of cohesion according to the interestigness measure MGK. In this paper, we present, for the first time, a validation of this approach in the field of education, mainly in the computing curricula and the performance capabilities of students pursuing this curriculum in the Anglo-Saxon model.

Mots clés: Binary data; Quality measure MGK; Implicative cohesion; Statistical implicative analysis; Oriented classification

Domaines: Education; Intelligence artificielle [cs.AI]; Environnements Informatiques pour l'Apprentissage Humain; Apprentissage [cs.LG]; Probabilités [math.PR]; Statistiques [math.ST]; Informatique

Hery Frédéric Rakotomalala, Totohasina André. On Hierarchical Classification Implicative and Cohesive MGK-Based: Application on Analysis of the Computing Curricula and Students Abilities According the Anglo-Saxon Model. Fourth International Congress on Information and Communication Technology ICICT 2019, Feb 2019, London, United Kingdom. pp.83-90, ⟨10.1007/978-981-15-0637-6⟩. ⟨hal-02904486⟩