Criminal Combat: Crime Analysis and Prediction Using Machine Learning
Amar Shukla, Avita Katal, Saurav Raghuvanshi, Shivam Sharma
Abstract
Crime is one of the most critical issues that the entire world is facing nowadays. The rate of crime should be minimized by using different techniques of machine learning in order to safeguard the global community from getting trapped into the activities of the criminals or anti-social elements. The paper identifies the crime patterns by utilizing the different mathematical and statistical models to forecast the probability of the crime. Crime datasets of the State of North Carolina have been used for this purpose. The paper aims to check different statistical parameters and Figure out the most common factors that affect crime. The univariate and bivariate exploratory analysis is used for extracting the most dominant features. The Akaike Information Criteria (AIC) method is used to drop out unimportant attributes followed by testing of model through Mean Absolute Error (MAE), Median Squared Error (MSE) and Root Mean Squared Error (RMSE) techniques. The work done in the paper concludes that crime predictability and criminology can be very useful in eliminating menace of crime from our society. These mathematical and statistical models can aid us in the process of making our society a safer place to live in.