Analysis of a Complex Dataset Using the Combined MST and Auto-Contractive Map

The auto-contractive map neural network is applied to the London Metropolitan Police drug trafficking database to identify previously unknown relationships and depicts the new knowledge in the form of a minimal spanning tree. Each subset of data is shown along with the corresponding minimal spanning tree, and a detailed description is presented to aid the reader in correctly interpreting the structure. The knowledge attained using the auto-contractive map can assist police agencies in directing limited resources to geographical areas and individuals that have the greatest probability of success. Profiles are created based entirely on the neural network algorithms that are independent of any human subjectivity.