Balancing Optimality and Robustness in Resource Allocation Problems

摘要

We formally define a flexible robustness mechanism for resource allocation problems encoded as combinatorial auctions based on repairable solutions. Our notion of robustness is focused on resources that become unavailable once an allocation has been found. We extend previous works on encoding auctions to weighted Max-SAT and supermodels for the boolean satisfiability framework (SAT) with a mechanism that adds flexibility to the robust solutions.