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 $\backslash$theta $,3$ and $5.$ The model is that X has the probability function p$($x$|\backslash$theta $)$

depending on $\backslash$theta $,$ given by the following table