Crime prediction: pattern recognition and prediction of collusive bidding

Abstract

Bidding is an effective means to promote fair competition in the market and optimize the allocation of social resources. However, with the increasing incidence of bid-rigging crimes, public resource regulatory agencies and public security agencies are facing new challenges. On the premise of fully studying the current trend of China’s domestic collusive bidding crimes, this paper summarized the abnormal characteristics of four categories of companies suspected of bid rigging, and selected seven representative subcategories of data based on actual conditions, and set thresholds for these data. Then machine learning methods were utilized to build a pre-warning model for collusion bidding. The results show that the use of machine learning methods to predict the bid collusion can achieve a maximum accuracy of 95%. The research can not only improve the early warning, monitoring, analysis and judgment capabilities of the economic crime investigation department, but ensure the benign operation of bidding, and protect the sustained and rapid development of the social economy.