Neural Networks

Question 1

What is a neuron? What is an activation function and which are the most common?

Answer 1

8 / 3-4

Question 2

Describe a neural network (type of graph, weight function, layers, notation, how a value in the node is processed, depth, size, width )

Answer 2

8 / 5-7

Question 3

What is the architecture of a NN? What could be an hypothesis class of a NN?

Answer 3

8 / 11

Question 4

Describe just orally how you could find the weights of NN with just 1 hidden layer that tries to implement a Lipschitz function. (Shit question, but you know what I mean)

Answer 4

8 / 15-17

Question 5

Describe the matrix notation for a NN (matrixes v, w, a)

Answer 5

8 / 25-27

Question 6

Write down the forward propagation algorithm

Answer 6

8 / 28

Question 7

What is the iter to compute the weights of a NN? (hard question)

Answer 7

8 / 29-35

Question 8

Briefly, what the backpropagation algorithm does? Main idea of the pseudocode?

Answer 8

8 / 37