Segmentação do perfil de clientes inadimplentes utilizando ferramentas computacionais
Alternative title
Profilsegmentierung von säumigen kunden mithilfe von computertoolsProfile segmentation of defaulting customers using computational tools
Segmentación de perfiles de clientes morosos utilizando herramientas computacionales
कम्प्यूटेशनल टूल का उपयोग करने वाले चूककर्ता ग्राहकों का प्रोफ़ाइल विभाजन
使用計算工具對違約客戶進行剖析
Author
Date
2022-03-26Type
View/ Open
Access rights

Metadata
Show full item recordAbstract
Default is a term that is very present in the lives of Brazilians, since in 2021 Brazil had its highest average debt level in 11 years. In addition, credit collection and recovery areas sought to increasingly improve their methods of approaching the customer, always aiming at the best strategy and the pursuit of profit. From this, it is possible to see how many studies with a credit recovery approach are important to improve processes and provide preventive actions, being these objectives of this work. Throughout the text it is possible to understand the credit path or the credit collection and recovery process, understand how some computational tools work, succinctly absorb the resource theory in logistics in addition to seeing it in practice from a mixed approach. among other Python and R software. In those evaluated under construction, the pre-feasible model to segment the profile of customers into non-defaulting customers was shown to be viable, whether a digital profile or not.
A inadimplência é um termo muito presente na vida dos brasileiros, visto que em 2021 o Brasil teve o seu maior nível de endividamento médio em 11 anos. Diante disso, áreas de cobrança e recuperação de crédito buscam esforços para melhorar cada vez mais seus métodos de abordagem ao cliente, sempre visando a melhor estratégia e a busca pelo lucro. A partir disso, é possível ver o quanto estudos com abordagem em recuperação de crédito são importantes para melhorar processos e proporcionar ações preventivas, sendo esses os objetivos deste trabalho. No decorrer do texto é possível entender o caminho do crédito até o processo de cobrança e recuperação de crédito, entender como algumas ferramentas computacionais funcionam, absorver de forma sucinta a teoria de regressão logística além de vê-la na prática a partir de uma abordagem mista entre algoritmos de software Python e R. Nas análises realizadas, foi demonstrado ser viável a construção de um modelo preditivo para segmentar o perfil de clientes inadimplentes, sendo a variável resposta o cliente ter um perfil digital ou não.
How to cite this document
Keywords
Language
Sponsor
