Describe at least two reasonable stopping criteria that can be used when training a neural network.
  • Stop after some fixed number of iterations.
  • Stop when the sum-of-squares error on the training data is less than some constant.
  • Stop when the derivatives all become within some ε of zero.
  • Split the data into training data and test data, and stop when the error on the test data increases.

Valid HTML 4.0 Transitional