Crime analysis and classification, as applied in law enforcement and public safety planning, are very crucial. The more structured the data was on criminal activities, the more promising was machine learning techniques in understanding the dynamics in criminal activities and predicting their future trends. The Seattle Police Department has provided data to classify crime into major categories through supervised machine learning algorithms such as K-Nearest Neighbors (KNN), Gaussian Naive Bayes, and AdaBoost with Decision Trees. The aim of this study is to investigate the performance of these three models and compare their results in terms of which provides the best classification accuracy for the task.
Full project link: https://github.com/emonbd/Crime_Report_Seattle