A deep learning approach to small area estimation

The growing demand for accurate local estimates in data-limited contexts represents a significant challenge for traditional statistical methods, such as Small Area Estimation (SAE). These methods often present difficulties when working with high-dimensional datasets, limited observations, and integrating heterogeneous information sources, such as geospatial data. In response to these limitations, this research explores using TabNet, a deep learning model designed for tabular data, as a computational solution to improve the accuracy and reliability of estimates in SAE problems.

The main objective of this work is to obtain reliable estimates using TabNet and evaluate its ability to handle the complexities inherent to SAE, especially in situations where traditional methods do not produce satisfactory results. Finally, our proposal aims to significantly impact public policy by enhancing the accuracy and reliability of estimates, which is crucial for optimizing resource allocation and budget planning in public health. Improved precision will enable more effective strategic planning, ensuring better-targeted interventions for vulnerable communities. This study addresses the perception of artificial intelligence models, often called "black boxes" due to their lack of interpretability. By employing TabNet, a model that combines robustness with interpretability, we seek to demonstrate that AI can deliver accurate results and clear and understandable explanations of its decisions. This will facilitate its adoption as a reliable tool in complex studies, promoting the use of more transparent and accessible AI models in public health and beyond.