This project uses the criminality data from Mexico City and creates clusters to identify the optimal number of "clusters" for a drone swarm first response system. For more information about the concept of this system, please see the following presentation:
https://www.slideshare.net/OscarAlvarez186/red-seguridad-drones
1 - Data_exploration.ipynb - This notebook goes through the structure of the database providing information about which data is relevant to this analysis.
2 - Data_cleaning.ipynb - Clean the data and prepare it for the clustering algorithms.
3 - Clustering_kmeans.ipynb - Cluster the data using kmeans algorithm.
4 - Clustering_kmedians.ipynb - Cluster the data using kmedians algorithm.
5 - Clustering_fcm.ipynb - Cluster the data using fuzzy c means algorithm.
6 - Nearby_search_police_stations.ipynb - Uses google maps API nearby seach module to locate police stations.
7 - Centroid_adjustment_and_calculations.ipynb - Adjust the centroids to the closest police station, calculates response times and plot the results.
8 - Plotting_Folium.ipynb - Plots the location of each centroid (drone base) on a map. A vuisual reprentation of the clusters over Mexico city map can be found in the next Tableau worksheet: