1. Suppose by using Gradient Descent optimization technique, you try to find optimal values of ?? and ??, in order to minimize the convex function J(??, ??) = ½(?? + ??)².
a. Update ?? and ??, for 3 iterations according to Gradient Descent (GD) optimization technique, with initial value of ?? = 0, ?? = 1. Take learning parameter ? = 0.2.
b. Repeat the same step of (a) with learning parameter ? = 1.
c. Make necessary observation from the computation of (a) and (b) and justify the reasons for having different answers.
d. Analyse what could possibly happen if J(??, ??) is a non-convex function? Justify.
6+3+2+4=15
