Building blocks of deep learning
Download as PPTX, PDF1 like633 views
The document provides an overview of key concepts in deep learning including deep neural networks, forward propagation, activation functions, cost functions, regularization, gradient descent, optimization, and backpropagation. It notes that backpropagation uses the chain rule of calculus to calculate the gradient of the loss function with respect to network parameters.
Building blocks of deep learning
- 1. Building blocks of Deep Learning Keshan Sodimana
- 2. Today’s menu ● What is a deep neural network ● How forward propagation works ● What are activation, cost function and regularization ● Gradient descent and optimizations ● Backpropagation ● Let’s practice….
- 3. BUT WAIT... Do I need to have a good knowledge in Mathematics? I have done nothing on Maths since my ALs.
- 4. y = mx + c z = f(x) ∑f(x) dx/dy Do you remember?
- 5. What is a deep neural network
- 9. Cost function Our goal is to minimize the cost.
- 10. Gradient descent Finding the global minimum
- 11. Optimization
- 12. Backpropagation Spoiler Alert : This could be the hardest part to digest
- 13. Backpropagation Remember : back propagation is used to calculate the gradient of the loss function with respect to the parameters
- 14. Can you remember the chain rule in calculus?
- 15. Ok let’s go for a test run...
Editor's Notes
- #16: Note to myself: iPython notebook needs to be prepared