Closed f-dangel closed 4 years ago
Reformulate the quadratic test problem as empirical risk minimization with a sequential neural network and square loss, see the issue in my fork for more details.
quadratic
Important: This will change the loss function of the test problem by a constant scale.
Reformulate the
quadratictest problem as empirical risk minimization with a sequential neural network and square loss, see the issue in my fork for more details.Important: This will change the loss function of the test problem by a constant scale.