FORECASTING THE NUMBER OF CORRUPTION CRIMES IN KAZAKHSTAN: A MACHINE LEARNING APPROACH

A. Bitanov

Data confidence 70 medium

How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

This study aims to predict the number of corruption crimes in Kazakhstan using machine learning methods. The research is based on official monthly crime statistics collected from the Legal Statistics Portal, specifically the Report Form No. 3-K, which records corruption-related offenses since 2016 [3]. Three regression models were applied: k-Nearest Neighbors (kNN), Extreme Gradient Boosting (XGBoost), and Linear Regression. Model performance was assessed using Mean Absolute Error (MAE), Mean Squared Error (MSE), and R-squared (R²) score. The findings indicate that Linear Regression achieved the highest predictive accuracy (R² = 1.000), followed by XGBoost (R² = 0.9977) and kNN (R² = 0.9333). These results suggest that machine learning models can effectively forecast corruption crime trends. This study highlights the potential of machine learning in corruption crime prediction. Future research can explore additional predictive features, alternative machine learning models, and real-time data integration to enhance forecasting accuracy.