Attribute Selection for Maintenance Time Prediction in Railway Freight Cars: An Analysis Using Filter Methods

• Josemar Coelho Felix Federal University of Ouro Preto (UFOP) http://orcid.org/0000-0003-0220-2346
• Rodrigo César Pedrosa Silva Federal University of Ouro Preto (UFOP)
• Andrea Gomes Campos Bianchi Federal University of Ouro Preto (UFOP)

DOI: https://doi.org/10.5753/sbbd.2024.243228

Abstract

This study investigated filter methods for selecting relevant attributes in predicting railcar maintenance time, using a database from the company MRS Logística and its experts. The results highlighted the sensitivity of the algorithms in attribute selection, with the Groupfs method proving efficient in predictive accuracy, despite having a higher computational cost. However, all methods faced challenges in distinguishing between pertinent and irrelevant attributes. As next steps, the integration of multiple company databases is planned to incorporate data on human and material costs, aiming to further optimize the railway maintenance process.

Keywords: MRS Logística, Machine Learning, Filter Methods

References

Alyasiri, O. M., Cheah, Y. N., Abasi, A. K., & Al-Janabi, O. M. (2022). Wrapper and hybrid feature selection methods using metaheuristic algorithms for English text classification: A systematic review. IEEE Access, 10, 39833-39852.

Bommert, A., Sun, X., Bischl, B., Rahnenführer, J., & Lang, M. (2020). Benchmark for filter methods for feature selection in high-dimensional classification data. Computational Statistics & Data Analysis, 143, 106839.

Hou, C. K. J., & Behdinan, K. (2022). Dimensionality reduction in surrogate modeling: A review of combined methods. Data Science and Engineering, 7(4), 402-427.

Felix, J. C., Oliveira, V. M., & Silva, R. (2022, November). A Machine Learning with an Inlier/Outlier Separation Approach for the Prediction of Wagon Maintenance Times. In Anais do X Symposium on Knowledge Discovery, Mining and Learning (pp. 9-16). SBC.

Rao, P. V., & Baral, S. S. (2011). Attribute based specification, comparison and selection of feed stock for anaerobic digestion using MADM approach. Journal of Hazardous Materials, 186(2-3), 2009-2016

Silva, L. R., & Nascimento, D. C. (2023). Avaliando o Processo de Seleção de Características na Tarefa de Junção de Similaridade. In Anais do XXXVIII Simpósio Brasileiro de Bancos de Dados (pp. 348-353). SBC.

Parmezan, A. R. S., Lee, H. D., & Wu, F. C. (2017). Metalearning for choosing feature selection algorithms in data mining: Proposal of a new framework. Expert Systems with Applications, 75, 1-24.

Penha, G., Cardoso, T. N., da Silva, A. P. C., & Moro, M. M. (2016). Análise de métodos de Inferência Ecológica em dados de redes sociais. In Anais do XXXI Simpósio Brasileiro de Bancos de Dados (pp. 109-114). SBC.

Pilnenskiy, N., & Smetannikov, I. (2020). Feature selection algorithms as one of the python data analytical tools. Future Internet, 12(3), 54.