3. Let Y, be the gth observation at time t;, j = 1, ..., n. Recall that the kernel estimator of the regression curve m(t) is given by m(t) = _ _w;(t)Y;, with the weights wj(t) = KA(t; - t)/ E Ke(ti-t) and a kernel function K (.). Show that m(t; ) will lie between the minimum observation, Y(), and the maximum observation, Y(), for all i = 1, ..., n, i.e., Yo)