A natural way to avoid the injection of potentially dangerous or illicit products in a certain country is by means of protection, following a strict portof- entry inspection policy. A naive exhaustive manual inspection is the most secure policy. However, the number of within containers allows only to check a limited number of containers by day. As a consequence, a smart port-ofentry selection policy must trade cost of inspection with security, in order to fit into the dynamic operation of a port. We explore the design of port-of-entry container inspection policies with imperfect information (unavailable or untrusted data). Starting from an a-priori classification provided by port-of-entry customs operator, a combinatorial optimization problem is introduced. The goal is to match an a-priori container classification with a logically coherent one, subject to a given level of container inspection. Inspired in the related literature, a novel Multi-Tree committee is introduced in order to find a solution to the previous combinatorial problem. It combines the strength of binary decision trees and minimization of logical functions. The algorithm is easy-to-handle and useful for an on-line production. We highlight the effectiveness of our proposal, regarding real traces available from the port of Montevideo. The results show the capability to detect the most risky containers and its conservative nature, respecting any desired level of inspection.