Why do we drop the summation term in the derivative of the regularized cost function

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

I

qn_drop_sum
qn_drop_sum961×477 60.2 KB

It’s mentioned in the video but not really explained, can someone explain to me why that happens

2

Hi @rmwkwok Raymond,

I think you had replied to this doubt recently. Probably you can direct to the same post.

Regards
DP

3

It’s because in a partial derivative, each weight is considered separately, and all the other weights are considered as constants.

Since the partial derivative of a constant is zero, all the other “non-j” terms become zeros, so they evaporate from the equation for each w_j.