Analysis of Vancouver Crime and Census Data Using Various Machine Learning Algorithms

No ratings

Presented at EISIC 2019 by

Kyle Behiels

Andrew Park

Justin Song

Valerie Tsang

In recent years mass storage of criminal data has become a common practice for many law enforcement agenciesaround the world. As these data sets grow, so does the amount of potential knowledge that we can gain fromanalyzing these data. When combined with census data and the open crime data set of Vancouver (non-violent crimesfrom between 2003 and 2019), we have a unique opportunity to explore the spatio-temporal snapshot of crime rateand causation. The resulting data lends itself extremely well to the application of various machine learningalgorithms. In this paper, we employs various machine learning algorithms to show the attributes correlation withmost strongly with both high and low crime rates.