Stochastic Greedy Algorithm Is Still Good: Maximizing Submodular + Supermodular Functions

  • Sai Ji
  • Dachuan Xu
  • Min Li
  • Yishui WangEmail author
  • Dongmei Zhang
Conference paper
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 991)


In this paper, we consider the problem of maximizing the sum of a submodular and a supermodular (BP) function (both are non-negative) under cardinality constraint and p-system constraint respectively, which arises in many real-world applications such as data science, machine learning and artificial intelligence. Greedy algorithm is widely used to design an approximation algorithm. However, in many applications, evaluating the value of the objective function is expensive. In order to avoid a waste of time and money, we propose a Stochastic-Greedy (SG) algorithm, a Stochastic-Standard-Greedy (SSG) algorithm as well as a Random-Greedy (RG) for the monotone BP maximization problems under cardinality constraint, p-system constraint as well as the non-monotone BP maximization problems under cardinality constraint, respectively. The SSG algorithm also works well on the monotone BP maximization problems under cardinality constraint. Numerical experiments for the monotone BP maximization under cardinality constraint is made for comparing the SG algorithm and the SSG algorithm in the previous works. The results show that the guarantee of the SG algorithm is worse than the SSG algorithm, but the SG algorithm is faster than SSG algorithm, especially for the large-scale instances.


BP maximization Stochastic greedy Approximation algorithm 


  1. 1.
    Bai, W., Bilmes, J.A.: Greed is still good: maximizing monotone submodular+ supermodular functions (2018). arXiv preprint arXiv:1801.07413
  2. 2.
    Bian, A., Levy, K., Krause, A., Buhmann, J.M.: Continuous dr-submodular maximization: structure and algorithms. In: Advances in Neural Information Processing Systems, pp. 486–496 (2017)Google Scholar
  3. 3.
    Bian, A.A., Buhmann, J.M., Krause, A., Tschiatschek, S.: Guarantees for greedy maximization of non-submodular functions with applications (2017). arXiv preprint arXiv:1703.02100
  4. 4.
    Bogunovic, I., Zhao, J., Cevher, V.: Robust maximization of non-submodular objectives (2018). arXiv preprint arXiv:1802.07073
  5. 5.
    Buchbinder, N., Feldman, M., Naor, J.S., Schwartz, R.: Submodular maximization with cardinality constraints. In: Proceedings of the Twenty-Fifth Annual ACM-SIAM Symposium on Discrete Algorithms, pp. 1433–1452 (2014)Google Scholar
  6. 6.
    Chekuri, C., Vondrák, J., Zenklusen, R.: Submodular function maximization via the multilinear relaxation and contention resolution schemes. SIAM J. Comput. 43(6), 1831–1879 (2014)MathSciNetCrossRefGoogle Scholar
  7. 7.
    Conforti, M., Cornuéjols, G.: Submodular set functions, matroids and the greedy algorithm: tight worst-case bounds and some generalizations of the rado-edmonds theorem. Discret. Appl. Math. 7(3), 251–274 (1984)MathSciNetCrossRefGoogle Scholar
  8. 8.
    Epasto, A., Lattanzi, S., Vassilvitskii, S., Zadimoghaddam, M.: Submodular optimization over sliding windows. In: Proceedings of the 26th International Conference on World Wide Web, pp. 421–430 (2017)Google Scholar
  9. 9.
    Fisher, M.L., Nemhauser, G.L., Wolsey, L.A.: An analysis of approximations for maximizing submodular set functions - II. Polyhedral Combinatorics, pp. 73–87 (1978)Google Scholar
  10. 10.
    Iwata, S., Orlin, J.B.: A simple combinatorial algorithm for submodular function minimization. In: Proceedings of the Twentieth Annual ACM-SIAM Symposium on Discrete Algorithms, pp. 1230–1237 (2009)Google Scholar
  11. 11.
    Kawase, Y., Sumita, H., Fukunaga, T.: Submodular maximization with uncertain knapsack capacity. In: Latin American Symposium on Theoretical Informatics, pp. 653–668 (2018)CrossRefGoogle Scholar
  12. 12.
    Kohli, P., Kumar, M.P., Torr, P.H.: P\(^3\) & beyond: move making algorithms for solving higher order functions. IEEE Trans. Pattern Anal. Mach. Intell. 31(9), 1645–1656 (2009)CrossRefGoogle Scholar
  13. 13.
    Krause, A., Guestrin, C., Gupta, A., Kleinberg, J.: Near-optimal sensor placements: maximizing information while minimizing communication cost. In: Proceedings of the 5th International Conference on Information Processing in Sensor Networks, pp. 2–10 (2006)Google Scholar
  14. 14.
    Lin, H., Bilmes, J.: A class of submodular functions for document summarization, pp. 510–520 (2011)Google Scholar
  15. 15.
    Mirzasoleiman, B., Badanidiyuru, A., Karbasi, A., Vondrák, J., Krause, A.: Lazier than lazy greedy. In: AAAI, pp. 1812–1818 (2015)Google Scholar
  16. 16.
    Narasimhan, M., Bilmes, J.: Pac-learning bounded tree-width graphical models. In: Proceedings of the 20th Conference on Uncertainty in Artificial Intelligence, pp. 410–417 (2004)Google Scholar
  17. 17.
    Nemhauser, G.L., Wolsey, L.A., Fisher, M.L.: An analysis of approximations for maximizing submodular set functions - I. Math. Prog. 14(1), 265–294 (1978)MathSciNetCrossRefGoogle Scholar
  18. 18.
    Niazadeh, R., Roughgarden, T., Wang, J.R.: Optimal algorithms for continuous non-monotone submodular and dr-submodular maximization (2018). arXiv preprint arXiv:1805.09480
  19. 19.
    Qian, C., Yu, Y., Tang, K.: Approximation guarantees of stochastic greedy algorithms for subset selection. In: Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence IJCAI, pp. 1478–1484 (2018)Google Scholar
  20. 20.
    Schoenebeck, G., Tao, B.: Beyond worst-case (in) approximability of nonsubmodular influence maximization. In: International Conference on Web and Internet Economics, pp. 368–382 (2017)CrossRefGoogle Scholar
  21. 21.
    Sviridenko, M.: A note on maximizing a submodular set function subject to a knapsack constraint. Oper. Res. Lett. 32(1), 41–43 (2004)MathSciNetCrossRefGoogle Scholar
  22. 22.
    Wei, K., Iyer, R., Bilmes, J.: Submodularity in data subset selection and active learning. In: International Conference on Machine Learning, pp. 1954–1963 (2015)Google Scholar

Copyright information

© Springer Nature Switzerland AG 2020

Authors and Affiliations

  • Sai Ji
    • 1
  • Dachuan Xu
    • 1
  • Min Li
    • 2
  • Yishui Wang
    • 3
    Email author
  • Dongmei Zhang
    • 4
  1. 1.College of Applied Sciences, Beijing University of TechnologyBeijingPeople’s Republic of China
  2. 2.School of Mathematics and StatisticsShandong Normal UniversityJinanPeople’s Republic of China
  3. 3.Shenzhen Institutes of Advanced Technology, Chinese Academy of SciencesShenzhenPeople’s Republic of China
  4. 4.School of Computer Science and TechnologyShandong Jianzhu UniversityJinanPeople’s Republic of China

Personalised recommendations