In the lesson entitled “Numerical Approximation of Gradients” Andrew makes a reference to;
g(\theta)
from the previous “slide” but no slides are being presented and there is no mention of this function in the previous video.
There is a video lesson missing between this video lesson and the previous one “Weight Initialization for Deep Networks” which should be explaining what this function is.
I am paying a monthly fee for this course and this situation of missing content is unacceptable and unfair.
@Mubsi
can you please look into this issue.
@Deept_Prasad I have found the “slides” after much searching but there is no corresonding Coursera video lesson.
He explained it in the lecture itself. He’s just defining it as a new function. He could have called it fred(\theta) or barney(\theta) if he wanted to be creative. Then he shows that it is a good approximation of f'(\theta), the derivative of f(\theta).
Here’s that section of the transcript of the lecture with the relevant section highlighted:
Here is the problem;
There is no “…previous slide…”. And there is no mention of g(\theta) in any of the previous video lessons for this week’s course.
Ok, I will suggest that they remove that quote, but (as I pointed out) he did explain what was going on here.
But in the past, it has proved pretty hard to get them to edit the videos.
1 Like
And you’ll notice that he’s saying exactly what I pointed out: that g(\theta) is the derivative of f(\theta), even if the previous slide did not include the notation g(\theta).
If f(\theta) = \theta^3, then f'(\theta) = 3 * \theta^2.
1 Like
And this is covered in the course slides:
So at some point in the past, that must have actually been included in the lectures.
Please indicate where in the transcript for the video lesson entitled “Numerical Approximation of Gradients” Andrew states that g(\theta) is the derivative of f(\theta) as I cannot find it.
You’re right that he doesn’t explicitly say that in the lecture. But I have explained why it is obvious that is what he meant.
If you want to cancel your subscription and ask for a refund, be my guest. Please realize that this is not going to get fixed anytime soon. These lectures have been the way they are since roughly 2017.
So you are in agreement with me that there is missing content from Week 1 of Course 2?
Yes, but it is an extremely minor issue, which can be easily explained, as I have done above.
I think it is important that course content should clear, complete and unambiguous, especially to students with little or no calculus knowledge.
