Underfitting / Overfitting

This is a very common problem that we face in the Machine Learning world.

Overfitting means your model works very well of the training set, but terribly on the test set.

We solve Overfitting through Regularization.