Presentation
Driven by the industrial sector and founded historically on data analysis and learning knowledge, extraction from data is a complex and interactive process. In order to supply useful knowledge from a mass of data methods and technologies of data, bases must be adapted to new constraints.In this course we examine the main algorithms of data search their strengths and their limitations. Evaluation of the quality of knowledge produced will be studied in depth, particularly in the context of non-supervised learning.
Duration:
63h
Organization
Scheduled activities
- C1 (3h) Introduction à la fouille de données (et à l'U.V.), données, méthodologie
- TP1 (3h) Démarche et méthodologie pour la fouille de données
- C2 (3h) Apprentissage non supervisé (clustering)
- TP2 (3h) Clustering
- C3 (3h) Apprentissage non supervisé (motifs fréquents et règles d'association)
- TP3 (3h) Règles d'association
- TDA1 (3h) Initiation du projet
- C4 (3h) Apprentissage supervisé (arbres, comparaison classifieurs, ensembles)
- TDA2 (3h) Projet
- TP4 (3h) Arbres de décision, comparaison classifieurs, ensembles
- C5 (1h30) Apprentissage supervisé (support vector machine)
- TP5 (1h30) Support vector machine
- C6 (3h) Apprentissage supervisé/non supervisé (méthodes connexionnistes)
- C7 (3h) Text mining
- C8 (3h) Text mining
- TP6 (3h) Text mining
- TDA3 (3h) Projet
- TDA4 (3h) Projet
- C9 (3h) Etude de cas de A à Z
- TP10 (3h) Etude de cas de A à Z
- TDA5 (3h) Projet
- Soutenances (1) (3h) Soutenances projets (1)
- Soutenances (2) (3h) Soutenances projets (2)
- Examen (3h) Examen
Team
|