Anais Da XIII Escola Regional De Alto Desempenho De São Paulo (ERAD-SP 2022) 2022
DOI: 10.5753/eradsp.2022.222234
|View full text |Cite
|
Sign up to set email alerts
|

Algoritmos de agrupamento aplicados à detecção de fraudes

Abstract: Em um contexto tecnológico, em que dados são gerados de maneira exponencial, as análises financeiras tem se tornado gradativamente mais importantes para evitar grandes perdas devido às fraudes. Neste trabalho, busca-se a segmentação das transações em grupos, por meio de técnicas de agrupamento, com base na existência de padrões distintos entre transações financeiras legítimas e ilegais. Para isto, algoritmos foram testados e comparados em relação ao desempenho, validação do agrupamento, interpretação e compree… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1

Citation Types

0
1
0

Year Published

2023
2023
2023
2023

Publication Types

Select...
1

Relationship

0
1

Authors

Journals

citations
Cited by 1 publication
(1 citation statement)
references
References 6 publications
0
1
0
Order By: Relevance
“…Dunn's Index measures the distance between the closest clusters relative to the average size of the clusters [Jain e Dubes 1988], used to identify optimal cluster compression while maintaining proper separation between clusters. Calinski-Harabasz is a measure of the density and separability between groups, used to pinpoint well-defined and densely-packed clusters with clear separations, while Davies-Bouldin measures the similarity between the group and its closest group [Furlanetto et al 2022], the metric was selected for its ability to highlight distinct and well-separated clusters.…”
Section: Resultsmentioning
confidence: 99%
“…Dunn's Index measures the distance between the closest clusters relative to the average size of the clusters [Jain e Dubes 1988], used to identify optimal cluster compression while maintaining proper separation between clusters. Calinski-Harabasz is a measure of the density and separability between groups, used to pinpoint well-defined and densely-packed clusters with clear separations, while Davies-Bouldin measures the similarity between the group and its closest group [Furlanetto et al 2022], the metric was selected for its ability to highlight distinct and well-separated clusters.…”
Section: Resultsmentioning
confidence: 99%