7. Evaluation: You should always keep aside some data points from your original training set for testing. Here we evaluate how the NN predicts data points it's never been trained on. 8/10
Evaluation: Testing Neural Networks on Unseen Data Points
By
–