1. Consider a set of training data points {(xi,yi)}i=1m sampled from a nonlinear curve y=f(x). Our goal is to find the function f from these given data points. Assume that you model f using the hypothesis y^=h(x) and define it as h(x)=a0x+a1x3+a2x5+a3x7. Design an algorithm to find the optimal hypothesis h that minimizes the empirical risk. [8 Marks]