Model 6
Deep learning model using BI-LSTM Layers
Autor | Eduardo Correa Araujo |
---|---|
Doença | Dengue |
Nível Administrativo | Município |
Resolução Temporal | Semana |
Linguagem de Implementação | Python |
Repositório | eduardocorrearaujo/forecast_models_dengue.git |
Descrição:
In this repo, in the path /models/neuralnetworks it's implemented a deep learning model with three
lstm layers (the first one is bidirectional) interspersed with dropout layers and a dense layer in the output.
This model computes the number of new cases in the next four weeks
based on the last four weeks of data (cases and climate variables).
The confidence interval of the predictions is computed using dropout and making multiple predictions to compute
the ci of them.