Multi-Objective Maximization of Monotone Submodular Functions with Cardinality Constraint

We consider the problem of multi-objective maximization of monotone submodular functions subject to cardinality constraint, often formulated as . While it is widely known that greedy methods work well for a single objective, the problem becomes much harder with multiple objectives. In fact, Krause et al.\ (2008) showed that when the number of objectives grows as the cardinality i.e., , the problem is inapproximable (unless ). On the other hand, when is constant Chekuri et al.\ (2010) showed a randomized approximation with runtime (number of queries to function oracle) . %In fact, the result of Chekuri et al.\ (2010) is for the far more general case of matroid constant. We focus on finding a fast and practical algorithm that has (asymptotic) approximation guarantees even when is super constant. We first modify the algorithm of Chekuri et al.\ (2010) to achieve a approximation for . This demonstrates a steep transition from constant factor approximability to inapproximability around . Then using Multiplicative-Weight-Updates (MWU), we find a much faster time asymptotic approximation. While the above results are all randomized, we also give a simple deterministic approximation with runtime . Finally, we run synthetic experiments using Kronecker graphs and find that our MWU inspired heuristic outperforms existing heuristics.
View on arXiv