Marco Antonio Lopes
Master's – Machine learning application in public frauds
Advisor: Profª. Drª. Alessandra de Ávila Montini
Comission: Profs. Drs. Adolpho Pimazoni Cantón, Vitor Hugo Louzada Patrício and Natália Cordeiro Zaniboni
Class: 215, FEA-5
In recent years, world governments have been participating in joint efforts to increase the availability of government data for their citizens, such as the creation of the Brazilian Open Data Portal. With more data available to the average citizen, several analyzes become possible, for example, the identification of fraud in public bids. One way to identify patterns in these scams is to use machine learning. These techniques can aid in the analysis of problems with unbalanced data, where the occurrence of the event being studied is very rare, as is the case of frauds. An example of this type of problem that will be the subject of the study of this dissertation is the detection of fraud in public systems. Thus, this work uses public data for the identification of frauds to the public patrimony.
*Abstract provided by the author