Title: Analyzing the risk of mass shootings in the United States
Abstract: Gun violence leads to tens of thousands of fatalities in the United States each year. Although a very small proportion of gun violence in the United States is due to mass shootings, the number of mass shootings in the United States represents a large proportion of global mass shootings. The number of mass shootings seems to be increasing in the United States. Since mass shootings take many innocent lives in the United States and generate significant media attention and public debate, it is necessary to understand the risks of mass shootings, analyze how the risk is changing over time, forecast the risk of mass shootings, and assess the risk for specific locations. This research models the annual trend in mass shootings by using several different models, to include Poisson regression, change-point models, a time series model, and a hybrid of a time series and neural network. The analysis of the fitted results provides useful insight into whether the historical frequency of mass shootings has changed. Several different change-point models are analyzed to ascertain which change-point model provides the best fit to the frequency of mass shootings. The forecasting performance of three forecasting models, a change-point model, a time series model, and a hybrid of a time series model with an artificial neural network model are compared. Each model is applied to forecast the frequency of mass shootings. Comparing among results from these models reveals advantages and disadvantages of each model when forecasting rare events such as mass shootings. The insights generated from the comparison are beneficial for selecting the best model and accurately estimating the risk of mass shootings in the United States. Quantifying the risk of mass shootings in the United States provides a good picture of risk at the national level but further analysis is needed to understand the risk at a state and local level. Quantifying the risk of a mass shooting at specific locations provides a greater understanding of the chance of a mass shooting at a school, house of worship, or even a workplace. A scientifically rigorous and evidence-based method is proposed to calculate the probability that a mass shooting occurs in a state or a specific location. Quantifying the risk of mass shootings at different locations in different areas can provide insight into which areas should be protected against mass shootings.