Abstract
We consider the problem of maximizing a non-negative monotone submodular function subject to a knapsack constraint, which is also known as the Budgeted Submodular Maximization (BSM) problem. Sviridenko (Operat Res Lett 32:41–43, 2004) showed that by guessing 3 appropriate elements of an optimal solution, and then executing a greedy algorithm, one can obtain the optimal approximation ratio of \(\alpha =1-{1}/{e} \approx 0.632\) for BSM. However, the need to guess (by enumeration) 3 elements makes the algorithm of Sviridenko (Operat Res Lett 32:41–43, 2004) impractical as it leads to a time complexity of roughly \(O(n^5)\) (this time complexity can be slightly improved using the thresholding technique of Badanidiyuru & Vondrák (in: SODA, 1497–1514, 2014) but only to roughly \(O(n^4)\)). Our main results in this paper show that fewer guesses suffice. Specifically, by making only 2 guesses (and using the thresholding technique of Badanidiyuru & Vondrák (in: SODA, 1497–1514, 2014), we get the same optimal approximation ratio of \(\alpha \) with an improved time complexity of roughly \(O(n^3)\). Furthermore, by making only a single guess, we get an almost as good approximation ratio of \(0.6174 > 0.9767\alpha \) in roughly \(O(n^2)\) time. Prior to our work, the only approximation algorithms that were known to obtain an approximation ratio close to \(\alpha \) for BSM were the algorithm of Sviridenko (Operat Res Lett 32:41–43, 2004) and an algorithm of Ene & Nguyen (in: ICALP, 53:1–53:12, 2019) that achieves \((\alpha -\varepsilon )\)-approximation. However, the algorithm of Ene & Nguyen (in: ICALP, 53:1–53:12, 2019) requires \({(1/\varepsilon )}^{O(1/\varepsilon ^4)}n\log ^2 n\) time, and hence, is of theoretical interest only since \({(1/\varepsilon )}^{O(1/\varepsilon ^4)}\) is huge even for moderate values of \(\varepsilon \). In contrast, all the algorithms we analyze are simple and parallelizable, which makes them good candidates for practical use. Recently, Tang et al. (in: Proc ACM Meas Anal Comput Syst, 5(1): 08:1–08:22, 2021) studied a simple greedy algorithm that already has a long research history, and proved that it admits an approximation ratio of at least 0.405 (without any guesses). The last part of this paper improves over the result of Tang et al. (in: Proc ACM Meas Anal Comput Syst, 5(1): 08:1–08:22, 2021) and shows that the approximation ratio of this algorithm is within the range [0.427, 0.462].
Similar content being viewed by others
Notes
We refer the reader to [26] for examples of such adaptations of an algorithm called Greedy \(^+\) which plays a central part in this paper and is introduced below.
Technically, Wolsey [25] proved this approximation ratio for another variant of Plain Greedy that outputs either the solution of Plain Greedy or a particular singleton calculated based on the execution of Plain Greedy. However, since Greedy considers all feasible singletons as possible solutions, it is at least as good as the algorithm of [25].
In the time complexity analysis, it is standard practice to assume that f can be evaluated on any given set in constant time.
We refer the reader to [26] for an example of the application of the thresholding technique to a greedy algorithm in the context of BSM. This technique can be applied to all the greedy algorithms considered in this paper in a similar way.
To be precise, Cohen & Katzir [3] considered a special case of BSM, and proved their result only for this special case. However, the proof extends to the general case.
Technically, this inequality holds for every value t in this range, except for maybe a finite number of points in which \(g'(t)\) is not defined. However, since g is continuous, we can safely ignore this technical issue.
References
Badanidiyuru, A., Vondrák, J.: Fast algorithms for maximizing submodular functions. In: SODA, pp. 1497–1514 (2014)
Călinescu, G., Chekuri, C., Pál, M., Vondrák, J.: Maximizing a monotone submodular function subject to a matroid constraint. SIAM J. Comput. 40(6), 1740–1766 (2011). https://doi.org/10.1137/080733991
Cohen, R., Katzir, L.: The generalized maximum coverage problem. Inf. Process. Lett. 108(1), 15–22 (2008)
De, A., Koley, P., Ganguly, N., Gomez-Rodriguez, M.: Regression under human assistance. In: AAAI, pp. 2611–2620 (2020). https://aaai.org/ojs/index.php/AAAI/article/view/5645
Elenberg, E.R., Dimakis, A.G., Feldman, M., Karbasi, A.: Streaming weak submodularity: Interpreting neural networks on the fly. In: NeurIPS, pp. 4047–4057 (2017)
Ene, A., Nguyen, H.L.: A nearly-linear time algorithm for submodular maximization with a knapsack constraint. ICALP, pp. 53:1–53:12 (2019)
Feldman, M., Naor, J., Schwartz, R., Ward, J.: Improved approximations for k-exchange systems - (extended abstract). In: Demetrescu, C., Halldórsson, M.M. (Eds.) ESA Lecture notes in computer science vol. 6942, pp. 784–798. Springer, Berlin (2011). https://doi.org/10.1007/978-3-642-23719-5_66
Feldman, M., Nutov, Z., Shoham, E.: Practical budgeted submodular maximization. CoRR arxiv:2007.04937 (2021)
Filmus, Y., Ward, J.: Monotone submodular maximization over a matroid via non-oblivious local search. SIAM J. Comput. 43(2), 514–542 (2014). https://doi.org/10.1137/130920277
Kazemi, E., Zadimoghaddam, M., Karbasi, A.: Scalable deletion-robust submodular maximization: Data summarization with privacy and fairness constraints. In: ICML, pp. 2549–2558 (2018)
Khuller, S., Moss, A., Naor, J.: The budgeted maximum coverage problem. Inform. Process. Lett. 70, 39–45 (1999)
Kulik, A., Schwartz, R., Shachnai, H.: A faster tight approximation for submodular maximization subject to a knapsack constraint. CoRR arxiv:2102.12879 (2021)
Kulik, A., Shachnai, H., Tamir, T.: Approximations for monotone and nonmonotone submodular maximization with knapsack constraints. Math. Oper. Res. 38(4), 729–739 (2013). https://doi.org/10.1287/moor.2013.0592
Lee, J., Sviridenko, M., Vondrák, J.: Submodular maximization over multiple matroids via generalized exchange properties. Math. Oper. Res. 35(4), 795–806 (2010). https://doi.org/10.1287/moor.1100.0463
Lei, Q., Wu, L., Chen, P., Dimakis, A., Dhillon, I.S., Witbrock, M.J.: Discrete adversarial attacks and submodular optimization with applications to text classification. In: MLSys, pp. 146–165 (2019). https://proceedings.mlsys.org/book/284.pdf
Libbrecht, M.W., Bilmes, J.A., Noble, W.S.: Choosing non-redundant representative subsets of protein sequence data sets using submodular optimization. Proteins: Struct., Funct., and Bioinfor. 86(4), 454–466 (2018)
Mirzasoleiman, B., Karbasi, A., Sarkar, R., Krause, A.: Distributed submodular maximization. J. Mach. Learn. Res. 17, 238:1-238:44 (2016)
Mitrovic, M., Morteza Zadimoghaddam, E.K., Karbasi, A.: Data summarization at scale: a two-stage submodular approach. In: ICML, pp. 3593–3602 (2018)
Nemhauser, G.L., Wolsey, L.A.: Best algorithms for approximating the maximum of a submodular set function. Math. Oper. Res. 3(3), 177–188 (1978). https://doi.org/10.1287/moor.3.3.177
Nemhauser, G.L., Wolsey, L.A., Fisher, M.L.: An analysis of approximations for maximizing submodular set functions-I. Math. Programm. 14, 265–294 (1978)
Salehi, M., Karbasi, A., Scheinost, D., Constable, R.T.: A submodular approach to create individualized parcellations of the human brain. In: MICCAI, pp. 478–485 (2017)
Sviridenko, M.: A note on maximizing a submodular set function subject to a knapsack constraint. Operat. Res. Lett. 32, 41–43 (2004)
Tang, J., Tang, X., Lim, A., Han, K., Li, C., Yuan, J.: Revisiting modified greedy algorithm for monotone submodular maximization with a knapsack constraint. Proc. ACM Meas. Anal. Comput. Syst. 5(1), 08:1-08:22 (2021). https://doi.org/10.1145/3447386
Ward, J.: A \((k+3)/2\)-approximation algorithm for monotone submodular k-set packing and general k-exchange systems. In: C. Dürr, T. Wilke (eds.) STACS, LIPIcs, vol. 14, pp. 42–53. Schloss Dagstuhl - Leibniz-Zentrum für Informatik (2012). https://doi.org/10.4230/LIPIcs.STACS.2012.42
Wolsey, L.A.: Maximising real-valued submodular functions: primal and dual heuristics for location problems. Math. Oper. Res. 7, 410–425 (1982)
Yaroslavtsev, G., Zhou, S., Avdiukhin, D.: “Bring your own greedy”+max: Near-optimal 1/2-approximations for submodular knapsack. In: AISTATS, pp. 3263–3274 (2020)
Funding
The research of Moran Feldman was supported in part by Israel Science Foundation (ISF) Grant No. 459/20.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
All authors certify that they have no affiliations with or involvement in any organization or entity with any financial interest or non-financial interest in the subject matter or materials discussed in this manuscript.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A: Missing Proofs of Sect. 4
In this section we give the proofs that have been omitted from Section 4.
Lemma 14
For every \(y \in [0, 1/2]\), there is a unique value \(z(y) \in [y, 1/2]\) satisfying the equation
Moreover, \(z'(y) = \frac{z(y) (1 - y - z(y))}{(1 - y)(z(y) + y)}\), z(y) is a non-decreasing function ofy, and \(z(y) \ge z(0) = 1/e\).
Proof
To see that the first part of the lemma holds for \(y \in (0, 1/2]\), note that as z(y) increases from y to 1/2, the left the hand side of the inequality defining z(y) continuously decreases from 0 to \(2y - 1\), and the right hand side of this inequality continuously increases to \(\ln \left( \frac{1/2}{1 - y}\right) \), which is less than 0 and at least \(- \ln (2 - 2y) \ge -[e^{\ln (2 - 2y)} - 1] = -(1 - 2y) = 2y - 1\). To see that the first part of the lemma holds for \(y = 0\) as well, note that by definition z(0) satisfies \(-1 = \ln (z(0))\), and \(1/e \in [0, 1/2]\) is the only number satisfying this equality.
Up to this point we have proved that the function z(y) is well defined. Our next objective is to show that it is also differentiable. To do that, consider the function \(G(y, z) = y / z - 1 - \ln \left( \frac{z}{1 - y}\right) \) and a point (y, z) obeying \(y \in [0, 1/2]\) and \(z = z(y)\). Clearly, \(G(y, z) = 0\), and the derivative dG(y, z)/dz obeys at this point
Hence, by the implicit function theorem, z(y) is a continuous differentiable function of y for this range of y.
Given the knowledge that the derivative of z(y) exists within the range \(y \in [0, 1/2]\), we can now calculate it by taking the derivative with respect to y of both sides of the inequality defining z(y). Doing so yields
and solving for \(z'(y)\) gives
where the inequality holds since the first part of the lemma shows that \(z(y) \in (0, 1/2] \subseteq (0,1-y]\). Note that, since the derivative \(z'(y)\) is non-negative, we get that z(y) is a non-decreasing function, as promised. \(\square \)
Lemma 15
\(\min \{p(x):x \in [0,1/3]\}=\frac{3-\ln 4}{4-\ln 4}\).
To make our calculations easier, it is useful to define \(y=\frac{x}{1-x}\), which implies \(x=\frac{y}{1+y}\) and \(1-x=\frac{1}{1+y}\), and therefore, also
We denote the rightmost side of the last equality by q(y). Since y goes exactly over all the values of the range [0, 1/2] when x grows from 0 to 1/3, \(\min \{p(x):x \in [0,1/3]\}=\min \{q(y):y \in [0,1/2]\}\). Thus, to prove the lemma it suffices to show that \(\min \{q(y):y \in [0,1/2]\} = \frac{3-\ln 4}{4-\ln 4}\), which we do in the rest of this proof.
Let us now use the shorthand \(z = z(y)\). Then,
where the second equality follow from Lemma 4. Since Lemma 4 guarantees that \(z \ge z(0) = 1/e\), the sign of \(q'(y)\) (within the range [0, 1/2]) is equal to the sign of \(2z + y - 1\), and the last expression is a (strictly) increasing function of y since \(z = z(y)\) is a non-decreasing function of y by Lemma 4. Hence, q(y) is a convex function within this range which takes its minimum value at the point \(y_0\) in which \(q'(y_0) = 0 \) (assuming there is such a point). In other words, to complete the proof of the lemma it remains to show that there is a point \(y_0 \in [0, 1/2]\) obeying \(q'(y_0) = 0\) and \(q(y_0) = \frac{3-\ln 4}{4-\ln 4}\).
We show that \(y_0 = \frac{1-\ln 2}{3-\ln 2} \in [0, 1/3]\) has the above mentioned properties. According to Lemma 4, \(z(y_0)\) is the sole value in \([y_0, 1/2]\) obeying the inequality
and one can verify that \(z(y_0) = (1 - y_0)/2 \in [y_0, 1/2]\) obeys this inequality. Hence,
and
\(\square \)
Appendix B: Code
This appendix includes the code used to get a lower bound on the expression
mentioned in Proposition 1. This code consists of two parts. The first part, given in Sect. 1, describes a structure used to represent non-negative rational numbers. The second part, given in Sect. 1, describes the main program, which uses the structure defined by the first part.
1.1 B.1 Structure Representing a Non-negative Rational Number
In this section we describe an immutable structure used to represent a non-negative rational number. The public member functions of this structure support various operations on such numbers. Some of these operations return the output one would expect, while others return a lower bound on this output. Member functions of the last kind are denoted by the prefix LB.
1.2 B.2 Main Program
In this section we give the main program used to evaluate the expression given in Proposition 1. This program uses the structure defined in Section 1.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Feldman, M., Nutov, Z. & Shoham, E. Practical Budgeted Submodular Maximization. Algorithmica 85, 1332–1371 (2023). https://doi.org/10.1007/s00453-022-01071-2
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00453-022-01071-2