Machine Learning Geo-spatial Framework for Crime Prediction: Based on Socioeconomic Factors

Moreira, Mary Cindrilla (2023) Machine Learning Geo-spatial Framework for Crime Prediction: Based on Socioeconomic Factors. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (3MB) | Preview
Preview	PDF (Configuration Manual) Download (7MB) | Preview

Abstract

Crime trends are the changes over time of criminal activity such as public order disturbances, offenses against the government, and controlled drug offenses. People involved in criminal activity are influenced by socioeconomic factors such as like income inequality, unemployment, population, poverty levels, access to education, housing quality, and GDP are closely associated with these changes. Identifying the most important socioeconomic factor that are contributing is a significant challenge since, there are numerous factors that affect. This research proposes a machine learning framework for crime prediction based on socioeconomic factors and Geo-spatial analysis of crime trends. The proposed framework combines a prediction model and geo-spatial classification model. The prediction model is implemented using of Random Forest. The Geo-Spatial (Geographic Information) research trains machine learning models for crime prediction using datasets that were scraped from the internet and the NYC government. Notable complaints are included in the primary data, which is complemented by socioeconomic statistics. Additionally included are spatial data on borough GDP, population trends, and police station locations. The research covers all five NYC boroughs and runs from 1950 to 2019. The Results are calculated by R-squared and Mean Squared Error, the study accurately forecasts crime trends. With an accuracy of 1450358.82 and an R-squared of 90%, the Random Forest model performs well. This study offers insightful information that could improve law enforcement activities, which could promote public safety initiatives. Its specific goals are to raise awareness in crime hotspots and improve socioeconomic variables that contribute to crime.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Stynes, Paul UNSPECIFIED
Uncontrolled Keywords:	Crime trends; Socioeconomic factors; Machine learning framework; Prediction models; Simple linear regression; Random Forest, Ordinary Least Square Regression, Classification model; KNN (K-Nearest Neighbors); Geospatial analysis
Subjects:	H Social Sciences > HV Social pathology. Social and public welfare > Criminology > Crimes and Offences H Social Sciences > HT Communities. Classes. Races Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	18 May 2025 13:51
Last Modified:	18 May 2025 13:51
URI:	https://norma.ncirl.ie/id/eprint/7571

Actions (login required)

View Item