Probabilistic Surrogate

Probabilistic surrogate models are computationally efficient approximations of complex, expensive-to-evaluate functions, primarily used to accelerate scientific discovery and engineering design. Current research focuses on improving the accuracy and efficiency of these surrogates, employing various architectures such as Gaussian processes, Bayesian neural networks, and neural implicit solvers, often within Bayesian optimization frameworks to guide the selection of optimal model parameters. This approach significantly impacts fields like materials science, analog circuit design, and safety-critical system validation by enabling faster exploration of high-dimensional parameter spaces and more reliable uncertainty quantification.