keyword
deep surrogate models
A deep surrogate model is a deep neural network trained to approximate the input-output behavior of a computationally expensive or complex system, simulation, or evaluation function. By learning patterns and nonlinear relationships from data produced by ground-truth processes, such as physical experiments, numerical simulations, or agent behavior evaluations, a deep surrogate model delivers rapid predictions at a fraction of the original computational cost. These models are commonly applied in surrogate-assisted optimization, scientific computing, engineering design, and reinforcement learning, enabling efficient iterative testing, exploration, and decision-making where repeated execution of the true underlying system would be prohibitively slow or resource-intensive.
1 item
