Definition 3 (Confidence)
Given a pattern P = P1, P2, ..., Pq and a DNA sequence database D, the confidence of P1P2 with respect to P1 is defined as Conf(P1P2, P1) = Sup(P1P2,D)/Sup(P1,D).
For example, the character "A" occurs 10 times and "AT" occurs 7 times, and in database D in Table 1, Conf(AT,A) = 7/10 = 0.7.