Data Science and Machine Learning (Theory and Projects) A to Z - Gradient Descent in RNN: Equations Exercise

The video tutorial discusses the parameter matrices W and the hidden state vector A, which has 10 elements. It explains the input vector X with 20 elements and how Y is computed using these vectors and matrices. The tutorial explores the necessary sizes and orders of matrices for computing Y and how to combine vectors and matrices to simplify the equation, resulting in a new vector Z.

What is the number of elements in the hidden state vector A?

• 10

• 15

• 5

• 20

How many elements does the input vector X have?

• 25

• 20

• 15

• 10

What is the equation used to compute Y?

• Y = W_a * X + W_x * A

• Y = W_a * X - W_x * A

• Y = W_a * A + W_x * X

• Y = W_x * A + W_a * X

What is the combined length of the vector when A and X are merged?

• 25

• 30

• 20

• 35

How can the equation for Y be simplified using a single vector and matrix multiplication?

• Y = W_a * Z

• Y = Z * W

• Y = W * Z

• Y = W_x * Z