A CLUSTERING-BASED APPROACH FOR IDENTIFYING GROUPS OF MUNICIPALITIES TO SUPPORT THE DIRECTION OF PUBLIC SECURITY POLICIES

The direction of public policies plays an important role in society as a whole, especially security, which, addition to being considered necessity for every citizen, is constitutionally guaranteed. This study presents the use unsupervised learning approach establishment clusters among municipalities State Pernambuco, Brazil, considering some types representative crimes, aiming direct actions prevent and fight crime order support policy makers. k-means algorithm was used main tool study, using software R 3.6.1, recommendations were directed each obtained clusters. To demonstrate direction, grouping with parameter k = 26 used, referring Security Integration Areas. results show that clustering provides greater effectiveness directing combat crime, given have greatest similarities are grouped same cluster.