| Nome do aluno | Andréa Jesus dos Santos |
| Título do trabalho | Transfer Learning between Recurrent Neural Networks using Heterogeneous Biosignals |
| Resumo do trabalho | The global health systems do not respond satisfactorily to individuals with neurological diseases. There are several resource constraints in this area: a reduced number of specialised health professionals, low quality of treatments offered, inefficient distribution of medications, and many other aspects. This scenario of the scarcity of care in human health directly affects the moment of diagnosis of diseases, initiation of treatment, and prescription of medications. It reduces the patient's quality of life with neurological, mental, or substance use problems. The World Health Organization revealed that neurological diseases account for 13% of all global illnesses in recent years. This research will investigate whether the exchange of information contained in heterogeneous electrical biosignals can support more accurate and timely diagnoses. This work proposes a model to evaluate whether the transfer of learning between recurrent neural networks that process heterogeneous electrical signals can increase the accuracy of predictive models. |
| Orientador | Marcos Ennes Barreto |
| Membro Titular 1 | Luciano Rebouças de Oliveira (PGCOMP) |
| Link para o curriculum lattes |
http://lattes.cnpq.br/ |
| Membro Titular 2 | Vinicius Gadis Ribeiro (UFRGS) |
| Link para o curriculum lattes |
http://lattes.cnpq.br/ |
| Suplente 1 | Vinicius Tavares Petrucci (PGCOMP) |
| Link para o curriculum lattes |
http://lattes.cnpq.br/ |
| Suplente 2 | Cristiano André da Costa (UNISINOS) |
| Link para o curriculum lattes |
http://lattes.cnpq.br/ |
| Data do exame | 24 Aug, 2022 |
| Horário do exame | 2:30 PM |
Data da Defesa:
24/08/2022 - 14:00
Tipo de Defesa:
Qualificação de Mestrado