##### Data Science For Dummies
Modeling the travel demand of criminal activity allows you to describe and predict the travel patterns of criminals so that law enforcement can use this information in tactical response planning. If you want to predict the most likely routes that criminals will take between the locations from where they start out and the locations where they actually commit the crimes, use crime travel modeling.

Travel demand modeling is the brainchild of civil engineers and was developed to facilitate improved transportation planning. Although you can use four different approaches in travel demand modeling — trip-based, integrated trip-based, tour-based, and activity schedule-based — the trip-based approach is most relevant to crime analysis.

A schematic that represents the travel demand model.

The trip-based approach is broken into the following four steps:

1. Trip generation. Model the trip production (the quantity of crime trips that originate in a zone of origination — a spatial region, like a neighborhood or subdivision) and the trip attractions (the quantity of crime trips that end in the zone of destination — the spatial region where the criminal act is executed).
2. Trip distribution. Incorporate a trip matrix — a matrix of rows and columns that covers a study area and depicts the patterns of trips across it— and a gravity model — a model that describes and predicts the locational flow of objects across space — to quantify the count of crime trips that occur between each zone of origination and each zone of destination.
3. Modal split. A modal split is the portion of travelers that uses particular trip paths across a study area. For travel demand modeling, you'd generate a count of the number of trips for each zone-of-origination/zone-of-destination pair that occurs via each available route. The choice between routes can be modeled statistical or mathematical.
4. Network assignment. Assign probability and predict the most likely routes that a criminal would take when traveling from a particular zone of origination to a particular zone of destination across the network of potential travel paths.

Lillian Pierson is the CEO of Data-Mania, where she supports data professionals in transforming into world-class leaders and entrepreneurs. She has trained well over one million individuals on the topics of AI and data science. Lillian has assisted global leaders in IT, government, media organizations, and nonprofits.