巴西登革热住院数据集（1999-2021）：月度数据周粒度分解

A dataset (v1.0.0) of dengue hospitalizations in Brazil is publicly released, available on Zenodo under DOI 10.5281/zenodo.18189192. Driven by the necessity to enhance the temporal granularity of originally monthly data for more effective training of AI models in epidemiological forecasting, this dataset harmonizes municipal-level dengue hospitalization time series across Brazil and disaggregates them to weekly resolution (epidemiological weeks). The dataset spans the period from 1999 to 2021, offering researchers and AI developers finer temporal resolution data to support the analysis of dengue transmission patterns and the development of predictive models. Transforming monthly aggregated data into weekly data allows for the capture of more nuanced epidemiological dynamics, which is crucial for short-term forecasting and early warning systems. The construction of this dataset involved standardizing and cleaning raw data from various sources and formats, ensuring data consistency and reliability. The included geographical information enables regional analysis, further revealing differences in dengue transmission across various cities and regions. The release of this dataset aims to foster open science, encourage broader collaboration to address public health challenges posed by dengue, and provide a high-quality data foundation for AI applications in public health. Data accessibility will help accelerate the innovation and deployment of dengue prediction algorithms, thereby improving the efficiency of disease surveillance and control.