| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 3.66 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
Money Laundering is a crucial topic in the financial sector, and it is necessary to have the right tools to combat this problem in our society. The Customer Risk Rating model is a technique that can impact and reduce the exposure of cases of ML. With all the increments in machine learning and data visualization, it is possible to improve the efficiency and effectiveness of these models. Studying how a specific tool can impact the calibration process of weighted models is crucial for keeping this anti-money laundering technique updated and adapted to each problem faced. The main objective of this study is to develop and provide dashboards with a comparison of results between two different visions, where one is based on real-time choices relative to the weights applied in the model. This solution will allow the presentation of a dynamic report with relevant statistics about the distributions and characteristics of the clients. Plus, it will improve and reduce the time of the decision-process of weights, since it will present the impact of each choice in real time. In this study a specific dataset was used, but the purpose of this solution is to be adaptive to new models or new risk factors in order to not be dependent on a specific model. The method of constructing the dashboards followed the Methodology presented in this thesis, and it took into consideration the principles of Data Visualization. Once the development of all the dashboards was complete, it was possible to evaluate the results obtained. This study can provide results that will guide further research regarding the calibration process of CRR Models.
Descrição
Dissertation presented as the partial requirement for obtaining a Master's degree in Information Management, specialization in Knowledge Management and Business Intelligence
Palavras-chave
Data Visualization Anti-Money Laundering Customer Risk Rating Models Dashboard Weights Parameters Power BI
