Solução do problema de marketing direto com canibalismo: métodos exatos e heurísticos

• Ano de defesa: 2020
• Autor(a) principal: Cruz, Greici da Rosa da
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; Brasil; Ciência da Computação; UFSM; Programa de Pós-Graduação em Ciência da Computação; Centro de Tecnologia
• Palavras-chave em Português: Marketing direto; Modelagem matemática; Canibalismo; Heurística; Direct marketing; Mathematical modeling; Cannibalism; Heuristic; CNPQ::CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://repositorio.ufsm.br/handle/1/22263

Today, the globalized world has a large volume of available data, which is essential for companies to create strategies to obtain advantages and stand out in an increasingly competitive market. Customers are increasingly demanding, making it necessary for the organizations to create methods to reach them in a targeted and interesting way, enabling the use of direct marketing. Direct marketing is a set of strategies that companies use to measure the best alternative to a given target audience, seeking to satisfy the need of each aiming to maximize the potentialities of organizations. In this context, this research solves the Direct Marketing Problem with Product Offering, using exact and heuristic methods, showing that the best heuristic algorithm in the literature is not robust. Also, the Cannibalism effect is inserted, which occurs when two products dispute the same market with each other and should not be offered simultaneously to the same customer. The insertion of the Cannibalism effect was performed in the exact and heuristic models that have been studied and compared. This problem shows that algorithms have a high difficulty in dealing with large instances, which are situations to close to reality, for this, it proposes the use of meta-heuristics. Tests were performed with 358 instances, without and with the effect of Cannibalism, obtaining 5864 results related to Model 1, Model 2, heuristic HR1, H-R2, Meta-heuristic Tabu GRASP, and dissimilarity and similarity criteria. Comparisons are made with all other articles existing in the literature, which robust results and better performance are obtained for the vast majority of problems. The files used in the search for the application of the exact and heuristic methods are available on GitHub (www.github.com), through the link https://github.com/greici/MARKETING-DIRETOCOM- CANIBALISMO-METODOS-EXATOS-E-HEURISTICOS, to assist the academic community, in new research with Direct Marketing Problem with Product Offering