Global Sensitivity Analysis - Saman Razavi

The JRC's Sensitivity Analysis group (SAMO) presents "A New Framework for Comprehensive, Efficient, and Robust Global Sensitivity Analysis", by Saman Razavi, University of Saskatchewan. Seminar at the European Commision Joint Research Centre (JRC) – Ispra – 2 May 2017. This presentation provides an overview of the theory and application of a new framework for Global Sensitivity Analysis (GSA), called Variogram Analysis of Response Surfaces (VARS). VARS utilizes the concepts of variograms and covariograms to characterize a spectrum of sensitivity-related information across the model factor space. VARS is a general framework with explicit theoretical relationships with variance-based (e.g., Sobol) and derivative-based (e.g, Morris) approaches to GSA, while being highly efficient and statistically robust. This presentation also discusses strategies for improved convergence and robustness of GSA, and to this end, introduces a sequential sampling algorithm, called Progressive Latin Hypercube Sampling (PLHS), which allows progressively increasing the sample size, while maintaining the required distributional properties. Saman Razavi received his PhD degree (2013) in civil engineering from the University of Waterloo, Ontario, and his MSc (2004) and BSc (2002) degrees in civil engineering from Amirkabir University and Iran University of Science and Technology in Iran. His research interests include environmental and water resources systems analysis, hydrologic modelling, single and multiple-objective optimization, sensitivity and uncertainty analysis, and climate change and impacts on hydrology and water resources. Seminar organiser: William BECKER Video recording, audio and video editing: Mayeul KAUFFMANN