Predicting Safe Parking Spaces: A Machine Learning Approach to Geospatial Urban and Crime Data

Abstract

This research aims to identify spatial and time patterns of theft in Manhattan, NY, to reveal urban factors that contribute to thefts from motor vehicles and to build a prediction model for thefts. Methods include time series and hot spot analysis, linear regression, elastic-net, Support vector machines SVM with radial and linear kernels, decision tree, bagged CART, random forest, and stochastic gradient boosting. Machine learning methods reveal that linear models perform better on our data (linear regression, elastic-net), specifying that a higher number of subway entrances, graffiti, and restaurants on streets contribute to higher theft rates from motor vehicles. Although the prediction model for thefts meets almost all assumptions (five of six), its accuracy is 77%, suggesting that there are other undiscovered factors making a contribution to the generation of thefts. As an output demonstrating final results, the application prototype for searching safer parking in Manhattan, NY based on the prediction model, has been developed.

Integration of Building Information Modelling (BIM) in Third-Generation Crime Prevention through Environmental Design (CPTED)

“…Target hardening will give a sense of pride and confidence to the residents in defending their homes [36]. These protection tools should be installed in house unit and carpark area as the lack of target hardening installed in carpark area provided more opportunities for thefts to steal the vehicles, especially motorcycles [37]. Although many literature are substantiating the effectiveness of target hardening in reducing crime rate in residential areas but being too confidence in the installed tools can give a false sense of full safety to the residents too [38].…”

Crime Prediction and Monitoring in Porto, Portugal, Using Machine Learning, Spatial and Text Analytics

“…Machine learning approaches have been widely applied in different fields, such as urban science, transport and pedestrian flow prediction, healthcare, biology, archeology, finance and even arts [38,39]. They have been used to monitor illegal activities [40,41] and to model and predict crime, with authors often comparing various methods [42][43][44][45][46].…”

Machine Learning Against Terrorism: How Big Data Collection and Analysis Influences the Privacy-Security Dilemma

“…One example is for recognising particular areas as hot spots for crime. Matijosaitiene et al ( 2019 ) have achieved very high accuracy in predicting car theft in urban areas of New York City. Camacho-Collados and Liberatore ( 2015 ) developed a decision support system that proposes when and where police patrols should be deployed based on data sets that capture the time and place crimes were committed in the past.…”

“…As more training data naturally implies a higher probability of achieving more reliable and accurate algorithms, the irregularity of data points can be amended in this way. In the case of localized crime prediction this can be very successful, whereby the set of data about crime used to train the algorithm is simply increased to a size that returns adequately accurate results (Matijosaitiene et al 2019 ). However, unlike in the case of predicting mundane crime, datasets related to (potential) terrorist attacks cannot be expanded so easily.…”