Mixed Robust/Average Submodular Partitioning: Fast Algorithms, Guarantees, and Applications

We study two mixed robust/average-case submodular partitioning problems that we collectively call Submodular Partitioning. These problems generalize both purely robust instances of the problem (namely max-min submodular fair allocation (SFA) Golovin (2005) and min-max submodular load balancing (SLB) Svitkina and Fleischer (2008)) and also generalize average-case instances (that is the submodular welfare problem (SWP) Vondrák (2008) and submodular multiway partition (SMP) Chekuri and Ene (2011a)). While the robust versions have been studied in the theory community Goemans et al. (2009); Golovin (2005); Khot and Ponnuswami (2007); Svitkina and Fleischer (2008); Vondrák (2008), existing work has focused on tight approximation guarantees, and the resultant algorithms are not, in general, scalable to very large real-world applications. This is in contrast to the average case, where most of the algorithms are scalable. In the present paper, we bridge this gap, by proposing several new algorithms (including those based on greedy, majorization-minimization, minorization-maximization, and relaxation algorithms) that not only scale to large sizes but that also achieve theoretical approximation guarantees close to the state-of-the-art, and in some cases achieve new tight bounds. We also provide new scalable algorithms that apply to additive combinations of the robust and average-case extreme objectives. We show that these problems have many applications in machine learning (ML). This includes: 1) data 1 ar X iv :1 51 0. 08 86 5v 2 [ cs .D S] 1 6 A ug 2 01 6 Wei, Iyer, Wang, Bai, Bilmes partitioning and load balancing for distributed machine algorithms on parallel machines; 2) data clustering; and 3) multi-label image segmentation with (only) Boolean submodular functions via pixel partitioning. We empirically demonstrate the efficacy of our algorithms on real-world problems involving data partitioning for distributed optimization of standard machine learning objectives (including both convex and deep neural network objectives), and also on purely unsupervised (i.e., no supervised or semi-supervised learning, and no interactive segmentation) image segmentation.