Data mining using nonlinear auto-associative artificial neural networks: The arrestee dataset

Abstract

Using the drug database, this chapter explores how to apply nonlinear auto-associative systems (non-supervised ANNs) to data analysis. The results of the analysis are presented in various minimal spanning trees and the graphs interpreted. While fascinating and informative graphs are created and discussed, these structures are the results of the application of mathematical algorithms to records of data, and it must be emphasized that the analysis depends directly on the quality of the data entry and that the results should be interpreted as a point of departure for anyone using these methods for investigative purposes. Specific ethnicities, for example, are linked to geographic areas, individuals to other individuals in the same or other areas, specific drug(s), the type of tactic used most successfully, etc. The individuals involved in criminal activity are further described by age, gender, association, the likely method of arrest, and by which group of officers.