Predictions
ID
Lang
Model name
Repository
Description
LSTM model for Infodengue Sprint
Predictions for 2023 in MS using the att architecture
CNNLSTM_DengueForecast_Cases_Climate_Data_Driven_Ensemble
Validation test 1 in MT
LSTM model with climate covariates (3 weeks ahead)
Prediction 3 weeks ahead using data up to epiweek 202501
infodengue_sprint_24_25_hybrid_CNN_LSTM_ensemble_model
infodengue_sprint_24_25_hybrid_CNN_LSTM_ensemble_model Prediction on CE_info_dengue_test_2
Chronos-Bolt
Validation set 3 for TO using Chronos-Bolt
Arima model (3 weeks ahead)
Prediction 3 weeks ahead using data up to epiweek 202531
infodengue_sprint_24_25_hybrid_CNN_LSTM_ensemble_model
infodengue_sprint_24_25_hybrid_CNN_LSTM_ensemble_model Prediction on DF_info_dengue_2024_2025
Random Forest model with uncertainty computed with conformal prediction
Forecast de novos casos para o geocode 2800308 entre 2022-01-01 e 2023-01-01 usando apenas os dados de todos as cidades clusterizadas com 2800308 como input
Chronos-Bolt
Prediction of Season 25-26 for TO using Chronos-Bolt
UERJ-SARIMAX-2025-2
Validation test 2 for UF=AL (UERJ-SARIMAX-2025-2)
Model 2 - Weekly and yearly (rw1) components
The model is founded on a structural decomposition designed for modeling counting series, employing a Poisson distribution. The log intensity is defined by the sum of weekly and yearly components, where the first one is defined as a AR(1) process and the last one follows a RW(1) process.
Dengue oracle - M2
Validation test 2 for sprint 2025 preds in PA
LSTM-RF model
LSTM-RF predictions for Mosqlimate Sprint 2025
Chronos-Bolt
Validation set 2 for CE using Chronos-Bolt
Model 1 - Weekly and yearly (iid) components
This upload represents the epidemic prediction for validation test 1 in the state BA.
Dengue oracle - M1
Validation test 1 for sprint 2025 preds in RN
Chronos-Bolt
Prediction of Season 25-26 for SC using Chronos-Bolt
Chronos-Bolt
Prediction of Season 25-26 for PA using Chronos-Bolt
Modelo fourier-gravidade
Prediction Validation test 1 in SC
Model 2 - Weekly and yearly (rw1) components
The model is founded on a structural decomposition designed for modeling counting series, employing a Poisson distribution. The log intensity is defined by the sum of weekly and yearly components, where the first one is defined as a AR(1) process and the last one follows a RW(1) process.
Chronos-Bolt
Prediction of Season 25-26 for MT using Chronos-Bolt
Cornell PEH - NegBinom Baseline model
Validation 2 (NegBinom Baseline model)
Chronos-Bolt
Prediction of Season 25-26 for BA using Chronos-Bolt
Chronos-Bolt
Prediction of Season 25-26 for GO using Chronos-Bolt
3054 predictions
https://api.mosqlimate.org/api/registry/predictions/?page=16&per_page=30&
Page 16 of 102
