Speaker: Alfredo Canziani
0:00:00 β Week 14 β Practicum
When training highly parametrised models such as deep neural networks there is a risk of overfitting to the training data. This leads to greater generalization error. To help reduce overfitting we can introduce regularization into our training, discouraging certain solutions to decrease the extent to which our models will fit to noise.
0:01:41 β Overfitting and regularization
0:18:11 β Model regularization (L2, L1, dropout, batch norm, and data augmentation)
0:49:30 β Visualizing Regularisation and Overfitting, Bayesian Neural Networks