Suppose we face a pattern recognition problem, where the data consists of a single set of pixels X (where there are 12 possible pixel patterns), and there are two possible patterns 0, "3" and "5". The model is that X has the probability function p(x|0) depending on 0, given by the following table. pixel pattern #x p(x 0): 2 3 4 5 6 7 8 9 10 11 12 0: "3" 0 0 .02 .03 .02 .03 .07 .15 .08 .14 .02 .44 0 .03 .01 .13 .08 .12 0 .08 .02 .20 .01 .32 (a) Find the best (in terms of type II error) test with significance level o = .10 for testing Ho : 0 = 3 vs. HA : 0 = 5.