Crime rate prediction using machine learning

Crime rate prediction is really important for helping the police find areas where a lot of crimes happen and stop things from occurring.This study uses computers to look at crime data from the National Crime Records Bureau for 19 big cities in India from 2014 to 2021.The National Crime Records Bureau has information on ten types of crimes like murder, kidnapping, crimes against women, cyber crimes and economic offences.To get the data ready we did some things to it like label encoding and picking the features.Then we tried five computer models.Support Vector Machine, K-Nearest Neighbor, Decision Tree, Random Forest and Neural Network.To see which one was best at predicting crime rates.The Random Forest model was the best.It got 93.20% of the predictions right when we tested it.We then used something called Flask to make the model useful for predicting crimes.The results show that using computers can really help the police understand and predict crimes, which can lead to decisions and use of resources.Crime rate prediction and the National Crime Records Bureau data can make a difference, in how the police do their job.