L rather than 1 used in superscripts for b and a

Course Q&A Machine Learning Specialization Advanced Learning Algorithms
1

In the slide detailing forward prop in bumpy (~ 4min 59sec https://www.coursera.org/learn/advanced-learning-algorithms/lecture/fZYiN/general-implementation-of-forward-propagation), where I was expecting to see 1 in superscripts for arrays a(1) and b, I see instead the character/letter “l”. Was this in error or am I missing something? I thought it may be “l” for any chosen layer, but it appears to be specific for the layer shown.

Also, where the W matrix definition is shown, I expected to see a comma between the two rows ( W = np.array([1, -3, 5][2, 4, -6]) ).

1 Like
2

You’re right. The character “l” in the superscripts (e.g., a[l], b[l]) refers to the general layer index and indicates that the operation or variable applies to any arbitrary layer l. For example, if the current layer is l=1, then a[0]=x. As for formatting the W matrix, the definition in NumPy should include commas to separate the rows, like this: W = np.array([[1, -3, 5], [2, 4, -6]])

1 Like
3

Thanks for clarifying. I think I got confused because sometimes the number and other times the character were used in the slide.