Crime Prediction & Prevention using Unsupervised Machine Learning

  • Prof. Rushali A.Deshmukh, Suraj Gaonkar, Vicky Khillare, Aniket Rokade


Crime rates are increasing day by day from the last 15 years in India due to various reasons.  Maharashtra is a state in India's western peninsular region in which crimes are increasing from the last few years. We are trying to predict the rate of various types of crimes, such as robbery, theft, murder, etc. We will generate crime hotspots on a map of Maharashtra based on the rate of a particular crime according to the location of the crime. We are collecting crime data of the past 15 years (2004- 2019) from various mediums like newspaper, Kaggle,, etc. Our algorithm is k-means, which is the simplest clustering algorithm. It is helpful to establish different relationships, such as the age of criminals and the type of crime they committed. We will also add the reasons behind the crime, such as money, drug abuse, etc. Our project will help police authority as well as the government for taking proper action for the reduction of the crime. It will be helpful for the investors for investment in a particular area according to the crime rate of that area as well as it is also helpful for tourists to decide which area is safe for tourism.