NORMA eResearch @NCI Library

Analysing Crime Patterns using Machine Learning: A case study in Chicago

-, Himanshi (2022) Analysing Crime Patterns using Machine Learning: A case study in Chicago. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
Preview
PDF (Master of Science)
Download (1MB) | Preview
[thumbnail of Configuration manual]
Preview
PDF (Configuration manual)
Download (2MB) | Preview

Abstract

Crime patterns analysis is systematic examination of high-volume crimes. Analysis of crime patterns is useful for predicting a crime before it occurs. Crime rates,hot spots, and types of crimes can all be reliably estimated from past patterns. Developing a machine learning model for predicting crime pattern with great accuracy is a challenge. By being able to forecast crimes based on time, location and so forth, law enforcement can provide useful information from a strategic standpoint. This research proposes the framework of Crime pattern analysis to predict the crime rate with an acceptable amount of accuracy using deep learning and supervised machine learning methods. The proposed framework uses deep learning methodology with traditional supervised machine learning approach. Chicago crime dataset for crime prediction starting from 2001 to December 2021 consisting of 62,59,111 crime records is used to train the deep learning for time series forecasting of crime rate using 24 classes of crime types. Feature engineering and data modelling are done to train six different time series model namely Weighted Moving Average, Exponential Moving Average, Simple Moving Average, Bidirectional LSTM, CNN-LSTM and Random Forest Regression. Result of six models is presented in this paper based on Root Mean Square Error (RMSE) and Mean Absolute Error (MAE). Bidirectional LSTM model outperforms all other models in terms of RMSE value. This paper forms a method to analyse crime patterns by doing the case study on Chicago Crime data and consequently creating deductive deductions with the aid of solid and trustworthy Machine Learning algorithms.

Item Type: Thesis (Masters)
Uncontrolled Keywords: Crimes; Crime Pattern Analysis and Forecast; Deep Learning
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
H Social Sciences > HV Social pathology. Social and public welfare > Criminology
Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Tamara Malone
Date Deposited: 26 Jan 2023 15:39
Last Modified: 03 Mar 2023 11:23
URI: https://norma.ncirl.ie/id/eprint/6133

Actions (login required)

View Item View Item