Recent questions tagged go_yt_dpp_ml_day1

Suppose the original linear regression gives the best weight as $w^*$.Now, we $\textbf{multiply every input feature by}$ $c$, but keep the output $y$ unchangedWhat will b...

Suppose the original loss is $\operatorname{Loss}(w)$, and we define:$$\operatorname{NewLoss}(w)=\log (\operatorname{Loss}(w))$$If $\log$ is a strictly increasing functio...

Consider the weighted linear regression loss:$$W R S S(w)=\sum_{i=1}^n r_i\left(w x^{(i)}-y^{(i)}\right)^2$$where $r_i$ is the weight (importance) of the $i$-th training ...