The quiz asks about 'approximation recursive formula' which hasn't been taught in the lessons

Course Q&A Math for Machine Learning Calculus for Machine Learning and Data Science
1

The final quiz asks about solving for approximation recursive formula (3 questions in the quiz) which hasn’t been taught in Coursera calculus course. How do I go about understanding the concept more? When I ask the AI coach it talks about approximation formula: X(k+1) = X(k) - f(Xk)/f’(Xk). This is different than the newton’s formula I’ve learnt where to determine the next step closer to 0, the formula is X(k+1) = X(k) - f’(Xk)/f’'(Xk).

In general, I’m a bit disappointed with the lack of coherency in lessons/quizzes/labs for the Calculus course. This has caused a lot of confusion and I’ve lost time in making sense of the material.

2

Indeed, there are issues throughout the M4ML course.

3

Would you have any suggestions on how I can learn approximation recursive formula concepts? Any articles/videos/links that would help? Thanks.

4

Gradient descent is covered in M4ML C2 Week 2. Newton’s method is also covered in the latter part of Week 3.

Gradient Descent is covered in DLS C1 and C2.

There are lots of lectures on YouTube about such concepts.

Here’s one of Prof Andrew’s lectures from DLS on YouTube.

Here’s the Stanford CS 229 version of similar material.

5

Thank you Paul. I understand Gradient Descent and also Newton’s method lessons taught in Coursera. The video links you’ve shared don’t specifically cover the approach I’d like to understand solving questions like “Using Newton’s method, find an approximation recursive formula for square root of 5”. I was able to find one from MIT that explicitly describes it, and it’s net-new topic that I believe should be explicitly taught.
I’ve walked through the lessons, and the quiz is a leap from what’s being discussed in the course.