Dive into Deep Learning (Study Group): Multilayer Perceptrons | Session 4

Dive into Deep Learning (Study Group): Multilayer Perceptrons | Session 4 In this fourth session of our "Dive into Deep Learning" study program, we will cover multilayer perceptrons and begin to discuss important techniques to train deep neural networks: - Overfitting & underfitting - Dropout - Weight decay - Numerical stability and initialization - Distribution shift In addition, you will learn how to make submissions of model predictions to Kaggle. Entire playlist: https://www.youtube.com/playlist?list=PLGSHbNsNO4ViFXawDmx-kEz7zGziOpNSb You can find more information about the deep learning study program and upcoming sessions here: https://github.com/dair-ai/d2l-study-group