Section Recap
Introduction
This short lesson summarizes key takeaways from section 42.
Objectives
You will be able to:
- Understand and explain what was covered in this section
- Understand and explain why this section will help you become a data scientist
Key Takeaways
The key takeaways from this section include:
- In deep learning a training, validation and test set are used when iteratively building the right deep networks
- Like traditional machine learning models, we need to watch out for the bias variance trade-off when building deep learning models
- Several regularization techniques can help us limit overfitting: L1 Regularization, L2 Regularization, Dropout Regularization,...
- Deep network training can be sped up by using normalized inputs
- Normalized inputs can also help mitigate a common issue of vanishing or exploding gradients
- You learned about gradient descent, but in deep learning some other optimization algorithms are introduced that work faster than gradient descent
- Examples of alternatives for gradient descent are: RMSprop, Adam, Gradient Descent with Momentum
- Hyperparameter tuning is of crucial important when working with deep learning models, as setting the parameters right can lead to great model improvements