Crime Data Analysis Using Juypter

Abstract

This project focuses on the analysis and prediction of
crime trends across various states and union territories in
India using machine learning techniques. The dataset
comprises crime-related statistics categorized by state,
district, and year. Initial data preprocessing steps include
handling missing values and removing duplicates to
ensure data quality. Exploratory Data Analysis (EDA) is
conducted through various visualizations to highlight
crime patterns, identify states with high and low crime
rates, and observe temporal trends in Indian Penal Code
(IPC) crimes.A machine learning model using Random
Forest Regressor is trained to predict the total number of
IPC crimes based on state, district, and year as input
features. Label encoding is used to convert categorical
variables into numeric format suitable for model training.
The model’s performance is evaluated using the Rsquared
metric, and predictions are visualized to compare
actual versus forecasted crime numbers.Furthermore, a
user interface component is incorporated, allowing users
to input a specific state, district, and year to receive a
crime forecast along with a safety classification (e.g.,
"Safest City", "Medium Safe City", or "Not Safe City").
This application can serve as a decision-support tool for
policymakers and law enforcement agencies to proactively
address crime trends