Defesa de Proposta de Dissertação: Davi Lotfi Lavor Navarro da Rocha
Data da publicação: 13 de junho de 2024 Categoria: Notícias, Proposta de Dissertação
Título: Gaussian Process Latent Variable Models for Anomaly Detection: Implementation and Practical Considerations
Data: 18/06/2024
Horário: 15h00
Local: Google Meet (https://meet.google.com/dsa-oobt-ywd)
Resumo:
Anomaly detection is critically important in finance, healthcare, and cybersecurity, enabling the identification of data outliers. Gaussian Process Latent Variable Models (GPLVMs) offer a versatile and probabilistic approach to the task of unsupervised learning. However, their application in anomaly detection frameworks is relatively unexplored. This research investigates the use of GPLVMs for anomaly detection with the goal to bridge this gap in the literature and to provide a foundation for future research. The proposed approach is rigorously evaluated with several benchmarks and compared with various alternative models. The results indicate that GPLVM ranks among the top methods, excelling in some particularly challenging scenarios, which are further detailed.
Banca examinadora:
- Prof. Dr. César Lincoln Cavalcante Mattos (MDCC/UFC – Orientador)
- Prof. Dr. João Paulo do Vale Madeiro (MDCC/UFC)
- Prof. Dr. João Paulo Pordeus Gomes (UFC)
