Recognition of Service Patterns Using Data Mining: Case Study

This research document presents a model for projections of services in different locations of each of the dependencies of the University of the Armed Forces - ESPE (UFAESPE). To meet this objective, data collected through surveys was used. Initially the problem that existed to project the type of service is detailed and the objectives to be achieved are established. Then theoretical basis of data mining with the techniques used is described and the KDD methodology that was applied in the investigation is detailed. Through the application of data mining techniques such as linear regression and neural networks, service projections were compared in each service unit or department located within the university campus. Subsequently, based on mining techniques and applied algorithms, it was possible to make projections of the type of service; helping in this way to plan the optimization of services for future years. Finally, the optimal model for determining the type of service is determined; a detailed report of the results obtained is presented and then conclusions reached.