Data Mining: Comparative study of predictive techniques

Today, a large number of companies are becoming aware of the wealth contained in their data, and are questioning the value of implementing techniques. Companies have access to ever more data. The sheer quantity of information available can make it very difficult to apprehend huge volumes of structured and unstructured data in order to implement company-wide improvement projects.This book focuses on 2 Datamining techniques (supervised and unsupervised) such as: decision trees, regression, neural networks and support vector machines (SVM)...we focus on the environment of use of each technique, the advantages,disadvantages and consequences of choosing one of these technical elements to extract hidden predictive information from large databases and how to implement each technique.Finally, the paper presented some valuable recommendations in the Dataminig field.

Autorentext

Professor Khalid BALARProfessor de ensino superior qualificadoEspecialidade: Business Intelligence e Modelagem EstatísticaCoordenador da Licença Profissional em E-Business e Gestão Digital.Faculdade de Ciências Jurídicas, Económicas e SociaisUniversidade Hassan II - Casablanca