Crime Prediction and Prevention using K-Means Clustering

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


Crimes as well as crime rates, are increasing from the last 10-15 years. Maharashtra is one of the developed states in western India in which crimes are growing from last 10-15 years. We propose the concept of k-means clustering to analyze the relationship between 2 things such as number of times particular crime takes place verses the age of criminal and so on. K-means clustering is an Unsupervised machine learning algorithm that is very much helpful in our system to analyze the relationship between available data and to identify the trend in the database. Our system successfully generates crime hotspots based on the locations of crime. It shows the location of crimes on the map. We have also included the reasons behind the crime such as money, drug abuse, personal reason, in our database. Our project is helpful for police authority to analyze the rate of different types of crime in a particular area as well as to analyze which age group is committing more crime. It will also helpful for foreign as well as Indian tourist to analyze which area is safe for tourism.