Modelo de Inteligencia artificial para el diagnóstico del dengue

The differential diagnosis of dengue from other arboviral diseases presents a significant clinical challenge in endemic regions. This study, beyond proposing an Artificial Intelligence (AI) model for dengue diagnosis, examines and prepares the environment of dengue datasets for training AI models, particularly machine learning models. This study evaluated the effectiveness of supervised machine learning techniques for predicting dengue infection using clinical and demographic data. Several parametric and non-parametric binary classification algorithms were evaluated through cross-validation and widely used performance metrics such as accuracy and F1-score. It was found that data quality affects model performance, as the binary model delivers better results in balanced datasets with better-processed data than in unbalanced datasets or those with noise in their records. Based on quantitative evidence, the study concludes that further study and experimentation with dengue datasets are needed to facilitate the training process for machine learning models.