Abstract | Keywords | Export | Availability | Bookmark

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

This book explores the potential of using machine learning methods to predict residential burglaries, with a focus on near-repeat phenomena. It investigates factors that contribute to accurate forecasting, utilizing data from 2010 to 2017 on residential burglaries, supplemented by geographic information about crime scenes. The study employs various machine learning techniques, including Random Forest, XGBoost, Support Vector Machines, Neural Networks, and Soft Voting, achieving a prediction precision of over 60%. The research also demonstrates for the first time the feasibility of making accurate forecasts for rural areas. This work is part of the 'BestMasters' series, highlighting top master's theses from universities in Germany, Austria, and Switzerland, and is intended for both practitioners and researchers in the fields of predictive policing, data science, and criminology.

Machine learning. --- Burglary investigation. --- Machine learning --- Burglary investigation