we need to say that the "actual Pareto frontier" may be unknowable in practice.
Hence, for pragmatic reasons, when comparing $N$ SBSE tools, it is common practice to use the union of the frontiers found by each tool $N_i$ as a surrogate for the "actual" frontier.
we need to say that the "actual Pareto frontier" may be unknowable in practice. Hence, for pragmatic reasons, when comparing $N$ SBSE tools, it is common practice to use the union of the frontiers found by each tool $N_i$ as a surrogate for the "actual" frontier.