Enhancing Environmental and Health Statistics through Artificial Intelligence: A Comparative Study of Imputation Techniques

In an increasingly globalized world, addressing health, environmental sustainability and social inequalities is crucial and requires an integrated approach involving national statistical offices. The latter is increasingly called upon to develop statistical frameworks to facilitate informed policy-making. However, incomplete or missing data in questionnaires or registers may compromise the accuracy and reliability of results. The main objective of this study is to assess the effectiveness of different imputation methods using machine learning (ML) and artificial intelligence (AI) techniques in dealing with missing data in social surveys. To this end, a comparative analysis of different imputation techniques has been carried out, based on real datasets from the Istat Multi-purpose Household Survey, where missing data are common. Preliminary results suggest that ML/AI-based imputation methods outperform traditional statistical techniques in terms of performance and robustness. The aim is to improve imputation techniques in official statistics to improve data quality on critical issues.