Implementing run_gradient_descent function for Multiple Linear Regression

Course Q&A Machine Learning Specialization Supervised ML: Regression and Classification
,
1

Hi Everyone,

As part of the week #2 optional lab C1_W2_Lab03_Feature_Scaling_and_Learning_Rate_Soln, I am trying to implement the run_gradient_descent on my own. Though it is not part of the course I think it is a good exercise for detailed understanding. Theoratically, I think I have the correct implementation however I am getting different results with this implementation vs the one in the function used in the lab.

I would highly appreciate your feedback. Thanks!

{0E856344-B5CF-4517-9C60-F76241D6DF43}
{0E856344-B5CF-4517-9C60-F76241D6DF43}937×418 11 KB

2

I have not tested your code. It would be useful if you post the results you are getting for the cost history vs. the iterations.

Start by looking at the shapes you are getting for the gradients (dj_dw and dj_db).

Then compare the gradients your code generates vs. the gradients from the code in the ungraded lab.