In this study, a new model based on data mining techniques is presented to predict police performance based on police indicators at the level of the Islamic Revolution. This model, emphasizing the instructions of the Supreme Leader, uses the indicators of God-centeredness, religious insight, morality, people-centeredness, justice-centeredness, excellence, and authority with kindness as criteria for evaluating police performance.
After preprocessing the data, the proposed model was used for quantitative features and trained using various machine learning algorithms including decision trees, support vector machines, random forests, and linear regression. Then, its efficiency in predicting police performance was analyzed using the criteria of mean square error, root mean square error, mean absolute error, and prediction accuracy.
The results showed that among the different algorithms, the linear regression algorithm has the best performance with an accuracy of 99 percent. This is while the worst performance is for the decision tree algorithm. This finding is also consistent with other indicators, and examining the results shows that linear regression has a lower prediction error compared to other algorithms.