Predicting and Preventing Crime: A Crime Prediction Model Using San Francisco Crime Data by Classification Techniques

Table 11

Accuracy, incorrectly classified instances, recall, and precision for Gradient Boosting Decision Trees on training and testing data.

Method	Accuracy (correctly classified instances)	Incorrectly classified instances	Recall	Precision
Gradient Boosting Decision Trees (training data)	99.44%	0.66%	98.89%	100%
Gradient Boosting Decision Trees (testing data)	99.75%	0.25%	100%	99.50%