Crime Rate Analysis and Prediction in Delhi Using Facebook Prophet

Abstract

As in every other developing country, crime has been a major problem in India. Several types of crimes occur in India, such as murder, rape, kidnapping, robbery and counterfeit. Such offenses have different statistical patterns, and they change over time. To find a substantial shift in pattern after a certain time period, this paper examines time series data of the crimes that occurred in Delhi. A predictive model is used to predict the future after the change occurs and to see if the data aligns with the forecast. For information visualization, a time series predictive algorithm developed by Facebook called Prophet is used. The prediction error is calculated using algorithms such as mean absolute percentage error (MAPE) and symmetric mean absolute percentage error (SMAPE), which are less than 5. The paper further proposes an inflection point algorithm that facilitates the prediction of trends between two dates. After applying the model to the data, it is observed that the graph assumes a downward trend in the years to come. Hence, there is an overall decrease in the crime rate.