Data Mining

Contenu du cours

Our world is getting more and more driven by data. The amounts of data collected is massively growing. Data Mining (a.k.a Knowledge Discovery in Database) is the science of extracting implicit, non-trivial, and potentially useful information from data. Three main tasks that data mining methods try to perform are: (1) to model the regularities in the data in the form of compact models and patterns, (2) to spot the irregularities, (3) to predict the future.

The course will cover the following topics:

Classification.
Model validation and data preparation
Clustering
Time series forecasting
Spatial and Spatiotemporal data mining
Frequent pattern and association rule mining.
Stream data mining
Outlier mining
Applications.

Objectifs (et/ou acquis d'apprentissages spécifiques)

This is a first course in Data Mining, with the following goals:

To introduce the fundamental concepts and techniques of data mining
To develop skills of using recent data mining software for solving practical problems
To establish the main characteristics and limitations of algorithms for addressing data mining tasks
To select the most appropriate combination of algorithms to solve a data mining problem
To develop and execute a data mining workflow on real-life datasets, and to solve a data-driven analysis problem
To identify promising business applications of data mining

Pré-requis et Co-requis

Connaissances et compétences pré-requises ou co-requises

This course assumes that participants have background on the following topics. There will be no formal check of this background at the inscription time. It is up to the participants to see and decide themselves.

Good knowledge of programming, with hands on experience in atleast one programming language
General knowledge of data structures, algorithms, and complexity analysis
General knowledge of databases and SQL

Méthodes d'enseignement et activités d'apprentissages

The course activities include lectures, exercise sessions, invited industry talk, industry visit if COVID-19 allows, and a group project. The lectures aim at building a fundamental understanding of popular data mining techniques strengths and limitations, as well as their associated computational complexity issues. The exercise sessions use Rapidminer as a tool for quickly prototyping data mining applications, suitable for engineers. The industry activities aim at connecting to the application world and get an idea about the state of practice. Finally the project is a chance to compile and further extend on the studied topics in a real big data mining application.

Références, bibliographie et lectures recommandées

Textbook:

Charu C. Aggarwal. Data Mining: The Textbook. Springer, 2015. (Available for download through Cible+).

Other useful readings:

David J. Hand, Heikki Mannila, Padhraic Smyth. Principles of Data Mining. MIT Press, 2001.
Delmater Rhonda, Hancock Monte. Data Mining Explained. Digital Press, 2001.
Pang-Ning Tan, Michael Steinbach, Vipin Kumar. Introduction to Data Mining. Pearson Education (Addison Wesley), 2006.
Rob J Hyndman and George Athanasopoulos. Forecasting: Principles and Practice (2nd ed). Monash University, Australia

Autres renseignements

Contacts

Lecturer: Mahmoud SAKR <mahmoud.sakr@ulb.be>
Assistant: Jean-Philippe HUBINONT <jean-philippe.hubinont@ulb.ac.be>

Campus

Solbosch

Evaluation

Méthode(s) d'évaluation

Examen écrit
Projet

Examen écrit

Projet

Construction de la note (en ce compris, la pondération des notes partielles)

The grade of the course will be split as follows:

Group project 40%
Written exam 60%

Langue(s) d'évaluation

anglais

Recherche