Gradient Descent algorithm \[ w := lr \frac{\partial Loss(w, b)}{\partial w} \] \[ \\ b := lr \frac{\partial Loss(w, b)}{\partial b} \] resources ml specialization ibm
