Ordenação de Algoritmos para Modelagem Preditiva de Churn: Analisando o Problema a Partir dos Métodos Sapevo-M e Vikor / Ordering Algorithms for Customer Churn Predictive Modeling: Analyzing the Problem Using Sapevo-M and Vikor Methods
Abstract
Modelagem preditiva para classificação de churn (abandono de cliente) é uma prática comum em empresas de diversos setores, no entanto, embora seja um tema vastamente explorado, escolher o classificador adequado pode ser uma tarefa árdua, dadas as particularidades de cada empresa e a variedade de algoritmos disponíveis. Portanto, a proposta deste artigo é a escolha de um algoritmo para modelagem preditiva de churn em uma startup brasileira. O processo decisório fez uso do método SAPEVO-M para obter os pesos dos critérios e definir uma medida de interpretabilidade para os algoritmos avaliados, e o método VIKOR foi utilizado para avaliar as alternativas. Após a aplicação dos métodos, os classificadores treinados com os algoritmos SVM (kernel radial) e Regressão Logística foram considerados os mais adequados para o modelo de negócio da empresa em questão.
Palavras-chave: SAPEVO-M; VIKOR. Tomada de Decisão. Modelagem Preditiva. Churn.
ABSTRACT
Predictive modeling for customer churn classification is a common practice among companies from different sectors, however even though it is a widely explored subject, choosing a suitable classifier may be a difficult task, given the particularities of each company and the variety of available algorithms. Therefore, the purpose of this article is to choice of an algorithm for customer churn predictive modeling in a Brazilian startup. The decision-making process used the SAPEVO-M method to obtain the weights of criteria and to define an interpretability measure for the evaluated algorithms, and the VIKOR method was used to evaluate the alternatives. After applying the methods, the classifiers trained with SVM (radial kernel) and Logistic Regression algorithms were considered the most suitable for the business model of the company in question.
Keywords: SAPEVO-M; VIKOR. Tomada de Decisão. Modelagem Preditiva. Churn.
References
REFERÊNCIAS
ABUABARA, Leila; PAUCAR-CACERES, Alberto; BURROWES-CROMWELL, Toni. Consumers’ values and behaviour in the Brazilian coffee-in-capsules market: promoting circular economy. International Journal of Production Research, v. 57, n. 23, p. 7269-7288, 2019. https://doi.org/10.1080/00207543.2019.1629664
CORREIA, Anderson Ribeiro et al. Modelo de negócio em uma comunidade agrícola: Aplicação de Soft Systems Methodology e Strategic Choice Approach. Modelo de negócio em uma comunidade agrícola: Aplicação de Soft Systems Methodology e Strategic Choice Approach, p. 1-388–416. https://doi.org/10.22533/at.ed.7881911075
CHECKLAND, Peter. Soft systems methodology: a thirty year retrospective. Systems research and behavioral science, v. 17, n. S1, p. S11-S58, 2000. https://doi.org/10.1002/1099-1743
DUCKSTEIN, Lucien; OPRICOVIC, Serafim. Multiobjective optimization in river basin development. Water resources research, v. 16, n. 1, p. 14-20, 1980. https://doi.org/10.1029/WR016i001p00014
GOMES, L. F. A. M.; MURY, Antonio-Roberto; GOMES, Carlos FS. Multicriteria ranking with ordinal data. Systems Analysis-Modelling-Simulation, v. 27, n. 2, p. 139-146, 1997.
HENNIG‐THURAU, Thorsten. Customer orientation of service employees: Its impact on customer satisfaction, commitment, and retention. International journal of service industry management, 2004. https://doi.org/10.1108/09564230410564939
HOSSIN, Mohammad; SULAIMAN, M. N. A review on evaluation metrics for data classification evaluations. International Journal of Data Mining & Knowledge Management Process, v. 5, n. 2, p. 1, 2015. https://doi.org/10.5121/ijdkp.2015.5201
KITCHNER, Karen Strohm. Cognition, metacognition, and epistemic cognition. Human development, v. 26, n. 4, p. 222-232, 1983. https://doi.org/10.1159/000272885
KUHN, Max; JOHNSON, Kjell. Measuring performance in classification models. In: Applied predictive modeling. Springer, New York, NY, 2013. p. 247-273. https://doi.org/10.1007/978-1-4614-6849-3_11
KUMAR, Pardeep et al. A benchmark to select data mining based classification algorithms for business intelligence and decision support systems. International Journal of Data Mining & Knowledge Management Process, p. 25–42, 2012. https://doi.org/10.5121/ijdkp.2012.2503
MARTENS, David et al. Performance of classification models from a user perspective. Decision Support Systems, v. 51, n. 4, p. 782-793, 2011. https://doi.org/10.1016/j.dss.2011.01.013
MILLER, Tim. Explanation in artificial intelligence: Insights from the social sciences. Artificial intelligence, v. 267, p. 1-38, 2019. https://doi.org/10.1016/j.artint.2018.07.007
MINGERS, John; ROSENHEAD, Jonathan. Problem structuring methods in action. European journal of operational research, v. 152, n. 3, p. 530-554, 2004. https://doi.org/10.1016/s0377-2217(03)00056-0
ONLINEOUTPUT. Vikor Software. 2017. Disponível em: http://www.onlineoutput.com/vikor-software/ Acesso em 01 fev. 2021.
PFEIFER, Phillip E. The optimal ratio of acquisition and retention costs. Journal of Targeting, Measurement and Analysis for Marketing, v. 13, n. 2, p. 179-188, 2005. https://doi.org/10.1057/palgrave.jt.5740142
SIMON, Herbert A. The structure of ill structured problems. Artificial intelligence, v. 4, n. 3-4, p. 181-201, 1973. https://doi.org/10.1016/0004-3702(73)90011-8
TEIXEIRA, L. F. H. ; SANTOS, M ; GOMES, C. F. S. . Plataforma on line SAPEVOWeb, 2020. Disponível em: www.sapevoweb.com Acesso em 01 fev. 2021.
___________________________________________ . PROPOSTA E IMPLEMENTAÇÃO EM PYTHON DO MÉTODO SIMPLE AGGREGATION OF PREFERENCES EXPRESSED BY ORDINAL VECTORS - MULTI DECISION MAKERS: UMA FERRAMENTA WEB SIMPLES E INTUITIVA PARA APOIO À DECISÃO MULTICRITÉRIO. In: XIX Simpósio de Pesquisa Operacional & Logística da Marinha, 2020, Rio de Janeiro. Simpósio de Pesquisa Operacional e Logística da Marinha - Publicação Online, 2019. p. 2326. https://doi.org/10.5151/spolm2019-168
VERBEKE, Wouter et al. New insights into churn prediction in the telecommunication sector: A profit driven data mining approach. European journal of operational research, v. 218, n. 1, p. 211-229, 2012. https://doi.org/10.1016/j.ejor.2011.09.031
DOI: http://dx.doi.org/10.12819/2021.18.5.8
Refbacks
- There are currently no refbacks.
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
ISSN 1806-6356 (Print) and 2317-2983 (Electronic)