Algorithmics for Data Mining
In the discipline of Data Mining, many technologies allow organizations to improve their processes on the basis of the analysis of existing data and the search for patterns in them. However, available success stories notwithstanding, is mandatory to acquire a consciousness of the limitations of these technologies: we will develop a study of the most usual algorithms and of their main parameters, so that the students become capable of identifying suitable tools for a given application. We will explain the theory and the practical usage of clusterers, associators, and classifiers so that the students acquire criteria to choose values for the many free parameters of each of these algorithms.
OBJECTIVES
-
To be aware of the theoretical and practical set of problems that constitute Data Mining, and to understand the main models and algorithms to tackle it: both at the conceptual level and at the level of their application through commercial tools, preferably open-source.
-
To acquire and demonstrate an ability to put to work the knowledge obtained in the autonomous, team-wise deployment of a practical data mining case, including a public presentation of the work developed.