We present two recent applications leading to very large-scale LP formulations. The first application involves finding optimal strategies for container inspection to intercept illicit weapons/materials hidden in ship cargoes. The other application involves embedding proteins into binary spaces such that Hamming distances are proportional to protein dissimilarities. For both problems, the developed large scale LP models provide a scalable and computationally efficient solution technique to derive the theoretically best possible solutions, improving on earlier approaches.

（連絡先：数理情報学　牧野　和久）