Congratulations, handbag X has surivived the UHbAD filter. Given this rigorous journey, the probability of handbag X being authentic is close to 1. 100% accuracy can never be determined without directly handling the bag (or buying from a licensed retail venue) - and even in these cases - you can never be exactly 100% certain. However, with this information, you can drastically reduce the possibility of that inevitable major headache that will insue after getting stuck with a counterfeit bag, and drastically increase the great buys you will encounter online.

---

U_bad = [QSS;Material;Label;Hardware;Brand_Specifics]

(i) X = handbag in filter
(ii) A = authentic
(iii) β = .07

X_authentic ⇔ -(P(A|U_x1:xn))log₂(P(A|U_x1:xn)) ≤ β
< -(P(A|U_x1:xn-1))log₂(P(A|U_x1:xn-1))
≤ -(P(A|U_x1:xn-2))log₂(P(A|U_x1:xn-2))
≤ -(P(A|U_x1:xn-3))log₂(P(A|U_x1:xn-3))
≤ -(P(A|U_x1:xn-4))log₂(P(A|U_x1:xn-4))
= X_fake

(*) If X passed a node, but with a % penalty, consider penalty (p) such that,

X_authentic ⇔ -(P(A|U_x1:xn) - p)log₂(P(A|U_x1:xn) - p) ≤ β < X_fake

---

:    :    :