Seminar 2. DeepONet 3 Types of Errors Approximation: the expressivity of the model. Optimization: how good the parameters are? Generalization: how good can the trained model match new samples? Uni...