Improvement of models for detecting collusion in Brazilian public tenders with statistical variables and explainable models

• Lucas D. Scoralick Federal Center for Technological Education Celso Suckow da Fonseca (CEFET/RJ)
• Diego N. Brandão Federal Center for Technological Education Celso Suckow da Fonseca (CEFET/RJ)
• Kele T. Belloze Federal Center for Technological Education Celso Suckow da Fonseca (CEFET/RJ) https://orcid.org/0000-0001-6257-2520

DOI: https://doi.org/10.5753/sbbd.2024.243170

Abstract

Collusions are secret agreements or combinations between two or more parties, usually to deceive or harm third parties. The practice of collusion in public tenders disrupts the market price balance, negatively impacting the costs and quality of public services. This study proposes a methodology to improve collusion classification models, using statistical variables combined with the analysis of explainable models to interpret results better. The results showed a significant prediction improvement of 1 to 4\%, except for the decision tree algorithm.

Keywords: collusion, bidding, statistic, improvement, SHAP, explainable models, machine learning, decision tree, logistic regression, naive bayes, support vector machine, SVM, random forest, multilayer perceptron, MLP, adaboost, quadratic discriminant analysis, QDA, ridge

References

Abrantes-Metz, R. M., Froeb, L. M., Geweke, J., and Taylor, C. T. (2006). A variance screen for collusion. International Journal of Industrial Organization, 24(3):467 – 486.

Busu, M. and Busu, C. (2021). Detecting bid-rigging in public procurement. a cluster analysis approach. Administrative Sciences, 11(1).

Du, Y. et al. (2023). Towards an explainable clinical decision support system for largefor-gestational-age births. PLoS ONE, 18(2 February).

García Rodríguez, M. J. et al. (2022). Collusion detection in public procurement auctions with machine learning algorithms. Automation in Construction, 133:104047.

Huber, M. and Imhof, D. (2019). Machine learning with screens for detecting bid-rigging cartels. International Journal of Industrial Organization, 65:277 – 301.

Imhof, D. and Wallimann, H. (2021). Detecting bid-rigging coalitions in different countries and auction formats. International Review of Law and Economics, 68.

Jiménez, J. L. and Perdiguero, J. (2012). Does rigidity of prices hide collusion? Review of Industrial Organization, 41(3):223 – 248.

Júnior, D. P. G., de Sousa Filho, G. F., and dos Anjos Formiga Cabral, L. (2023). Classificação de fraudes em licitações públicas através do agrupamento de empresas em conluios. Anais do XI Workshop de Computação Aplicada em Governo Eletrônico.

Nguyen, H. V. and Byeon, H. (2023). Prediction of parkinson’s disease depression using lime-based stacking ensemble model. Mathematics, 11(3).

Porter, R. H. and Zona, J. D. (1999). Ohio school milk markets: An analysis of bidding. RAND Journal of Economics, 30(2):263 – 288.

Silva, E., Fragoso, G., Rodrigues, L., Freitas, N., and Vinuto, T. (2022). Exposer: Framework para detecção de anomalias em licitações públicas. In Anais Estendidos do XXXVII Simpósio Brasileiro de Bancos de Dados, pages 77–82, Porto Alegre, RS, Brasil. SBC.

Silveira, D., de Moraes, L. B., Fiuza, E. P., and Cajueiro, D. O. (2023). Who are you? cartel detection using unlabeled data. International Journal of Industrial Organization, 88.

Silveira, D., Vasconcelos, S., Resende, M., and Cajueiro, D. O. (2022). Won’t get fooled again: A supervised machine learning approach for screening gasoline cartels. Energy Economics, 105.

Wallimann, H., Imhof, D., and Huber, M. (2023). A machine learning approach for flagging incomplete bid-rigging cartels. Computational Economics, 62(4):1669 – 1720.

Wang, Z. et al. (2023). Developing an explainable machine learning model to predict the mechanical ventilation duration of patients with ards in intensive care units. Heart and Lung, 58:74 – 81.