Universidade Nova de Lisboa

Universidade Nova de Lisboa

NOVA Information Management School

Data Mining

Code

100031

Academic unit

NOVA Information Management School

Credits

6.0

Teacher in charge

Weekly hours

45.0

Teaching language

Portuguese. If there are Erasmus students, classes will be taught in English

Objectives

In terms of skills, this discipline aims to stimulate the student to:
• the analysis and synthesis;
• The organization and planning;
• The writing and speaking in Portuguese ;
• Problem solving , partially structured ;
• The ability to make decisions ;
• Teamwork ;
• The ability to apply acquired knowledge in practice ;
• The ability to generate new ideas ( creativity ) ;
• Leadership ;
• Work independently ;

Prerequisites

 

Subject matter

Introduction to Data Mining;
Predictive models and descriptive models;
Inductive learning;
Methodology for Data Mining;

  • The process;
  • The definition of the problem;
  • Measuring the quality of the models;
Exploration and evaluation of data;
Visualization tools;
Preparation and pre-processing of data;
Descriptive models;
  • Market basket analysis
  • RFM analysis;
  • Clustering algorithms (K-Means);
  • Self-Organizing Maps;
  • Topics about segmentation databases;
Predictive Models
  • simple classifiers
  • Introduction to Bayesian classifiers
  • Classification based on instances
  • Drawing a learning system;
  • Classification Trees - DDT, Cart and C 4.5
  • Neural Networks - Layered with perceptron training by Backpropagation
  • Additional Topics on Predictive Modeling

Bibliography

Berry, M. and G., Linoff, Mastering Data Mining: The Art and Science of Customer Relationship Management. 2000, Brisbane: John Wiley & Sons.
Hand, D., Mannila, H., Smyth, P., ‘Principles of Data Mining’. MIT Press. 2001. ISBN 026208290X
Course Notes Enterprise MinerTM: Applying Data Mining Techniques, SAS Institute
Livro da disciplina

Teaching method

Lectures were theory is presented
Practical classes in computer rooms allowing students to apply the presented concepts.
Tutorial classes in which students must work autonomously,

Evaluation method

Continuous assessment - Test 1 (35%), Test 2 (35%), Project (30%)
2nd epoch - Exam (75%), Project (25%)

Courses