His research focuses on methods and models for informing a decision-making process. He has developed the Differential Importance Measures, inserted since 2002 in the “NASA Probabilistic Risk Assessment Procedures Guide for NASA Managers and Practitioners”, the d-importance measure, the finite change sensitivity indices and other methods aimed at improving the palette of tools for the sensitivity analysis of computer codes. He is working on new notions such as information density, on new feature importance measures for applications in machine learning, on mathematical methods for trend analysis and interaction quantification. For input-output mappings with multivariate responses, he is studying the extension of probabilistic sensitivity measures through the combination with optimal transport theory, a connection that leads to a novel and elegant framework. The applications of his methods range from business plans, to risk assessment models used for space risk analysis at NASA, to image datasets of artificial intelligence, to simulators used in hydrology for precipitation forecasting.