More data: A model can not memorize the training examples

Smaller/Simpler model: A less powerful, but a sufficient model in performance

Regularization: Can penalize large weights for overly complex models

For Neural Networks: Can try smaller and fewer layers

Not answering