Global Sensitivity

Global sensitivity analysis assesses the influence of input parameters on a system's output, moving beyond simple one-at-a-time perturbations to consider complex interactions. Current research focuses on developing efficient methods, such as variance-based approaches and smooth sensitivity techniques, to quantify these influences within diverse models, including Bayesian networks, rule lists, and contextual bandit algorithms. This work is crucial for improving the accuracy and privacy of machine learning models, as well as for enhancing the robustness and interpretability of complex systems across various scientific domains. Furthermore, understanding global sensitivity is vital for designing more reliable and efficient algorithms in applications ranging from risk assessment to personalized recommendations.