CMStatistics 2023: Start Registration
View Submission - CMStatistics
B0268
Title: Predicting the movement of anti state criminal gangs Authors:  Karthik Sriram - Indian Institute of Management India (India) [presenting]
Abstract: Many countries globally face the challenge of armed conflicts with anti-state criminal gangs. Unlike criminals associated with common crimes such as robbery or theft, anti-state criminal gangs are often driven by an ideology and deliberately strategise to debilitate the government by causing damage to national properties such as roads, bridges, factories, police facilities and terrorising citizens. Such gangs are constantly on the move to avoid being caught or neutralised by the police forces. It is therefore essential that the police anticipate the movements of the gangs to be proactive in tackling them rather than just reacting to their attacks. A novel statistical model is proposed to predict the movement of anti-state criminal gangs by systematically integrating information on their past movements, contextually important features such as forest density (modelled using satellite image data) or police camp locations, impacting their preference for some locations, and intelligence information. A Bayesian estimation procedure is given based on a particle filtering algorithm by exploiting the structure of the model, to dynamically estimate the parameters and generate predictions. Ideas are developed by considering the case of Naxalite criminal gangs that operate in India, using data obtained from the Indian police department.