Continuous DR-submodular Maximization: Structure and Algorithms

DR-submodular continuous functions are important objectives with wide real-world applications spanning MAP inference in determinantal point processes (DPPs), and mean-field inference for probabilistic submodular models, amongst others. DR-submodularity captures a subclass of non-convex functions that enables both exact minimization and approximate maximization in polynomial time. In this work we study the problem of maximizing non-monotone DR-submodular continuous functions under general down-closed convex constraints. We start by investigating geometric properties that underlie such objectives, e.g., a strong relation between (approximately) stationary points and global optimum is proved. These properties are then used to devise two optimization algorithms with provable guarantees. Concretely, we first devise a "two-phase" algorithm with $1/4$ approximation guarantee. This algorithm allows the use of existing methods for finding (approximately) stationary points as a subroutine, thus, harnessing recent progress in non-convex optimization. Then we present a non-monotone Frank-Wolfe variant with $1/e$ approximation guarantee and sublinear convergence rate. Finally, we extend our approach to a broader class of generalized DR-submodular continuous functions, which captures a wider spectrum of applications. Our theoretical findings are validated on synthetic and real-world problem instances.

[1]  Satoru Iwata,et al.  A combinatorial strongly polynomial algorithm for minimizing submodular functions , 2001, JACM.

[2]  Andreas Krause,et al.  Submodular Function Maximization , 2014, Tractability.

[3]  Jan Vondrák,et al.  On Multiplicative Weight Updates for Concave and Submodular Function Maximization , 2015, ITCS.

[4]  Zeyuan Allen Zhu,et al.  Variance Reduction for Faster Non-Convex Optimization , 2016, ICML.

[5]  Huy L. Nguyen,et al.  A Reduction for Optimizing Lattice Submodular Functions with Diminishing Returns , 2016, ArXiv.

[6]  Andreas Krause,et al.  From MAP to Marginals: Variational Inference in Bayesian Submodular Models , 2014, NIPS.

[7]  Donald M. Topkis,et al.  Minimizing a Submodular Function on a Lattice , 1978, Oper. Res..

[8]  Stefanie Jegelka,et al.  Robust Budget Allocation Via Continuous Submodular Functions , 2017, Applied Mathematics & Optimization.

[9]  Amin Karbasi,et al.  Gradient Methods for Submodular Maximization , 2017, NIPS.

[10]  Amnon Shashua,et al.  Nonnegative Sparse PCA , 2006, NIPS.

[11]  Evdokia Nikolova,et al.  Maximizing Non-Monotone DR-Submodular Functions with Cardinality Constraints , 2016, ArXiv.

[12]  Jan Vondrák,et al.  Maximizing a Submodular Set Function Subject to a Matroid Constraint (Extended Abstract) , 2007, IPCO.

[13]  Yuichi Yoshida,et al.  A Generalization of Submodular Cover via the Diminishing Return Property on the Integer Lattice , 2015, NIPS.

[14]  Jan Vondrák,et al.  Optimal approximation for the submodular welfare problem in the value oracle model , 2008, STOC.

[15]  S. Friedland,et al.  Submodular spectral functions of principal submatrices of a hermitian matrix, extensions and applications , 2010, 1007.3478.

[16]  Andrea Montanari,et al.  Non-Negative Principal Component Analysis: Message Passing Algorithms and Sharp Asymptotics , 2014, IEEE Transactions on Information Theory.

[17]  Francis Bach,et al.  Submodular functions: from discrete to continuous domains , 2015, Mathematical Programming.

[18]  László Lovász,et al.  Submodular functions and convexity , 1982, ISMP.

[19]  T. Motzkin,et al.  Maxima for Graphs and a New Proof of a Theorem of Turán , 1965, Canadian Journal of Mathematics.

[20]  Simon Lacoste-Julien,et al.  Convergence Rate of Frank-Wolfe for Non-Convex Objectives , 2016, ArXiv.

[21]  Saeed Ghadimi,et al.  Mini-batch stochastic approximation methods for nonconvex stochastic composite optimization , 2013, Mathematical Programming.

[22]  M. L. Fisher,et al.  An analysis of approximations for maximizing submodular set functions—I , 1978, Math. Program..

[23]  Suvrit Sra,et al.  Scalable nonconvex inexact proximal splitting , 2012, NIPS.

[24]  Ben Taskar,et al.  Near-Optimal MAP Inference for Determinantal Point Processes , 2012, NIPS.

[25]  Vijay K. Garg,et al.  Introduction to Lattice Theory with Computer Science Applications: Garg/Introduction to Lattice Theory with Computer Science Applications , 2015 .

[26]  Chandra Chekuri,et al.  Submodular function maximization via the multilinear relaxation and contention resolution schemes , 2011, STOC '11.

[27]  Jan Vondrák,et al.  Submodular maximization by simulated annealing , 2010, SODA '11.

[28]  Alexander J. Smola,et al.  Proximal Stochastic Methods for Nonsmooth Nonconvex Finite-Sum Optimization , 2016, NIPS.

[29]  Andreas Krause,et al.  Guaranteed Non-convex Optimization: Submodular Maximization over Continuous Domains , 2016, AISTATS.

[30]  I. Gijbels,et al.  Penalized likelihood regression for generalized linear models with non-quadratic penalties , 2011 .

[31]  Shinji Ito,et al.  Large-Scale Price Optimization via Network Flow , 2016, NIPS.

[32]  Maryam Fazel,et al.  Designing smoothing functions for improved worst-case competitive ratio in online optimization , 2016, NIPS.

[33]  Ben Taskar,et al.  Determinantal Point Processes for Machine Learning , 2012, Found. Trends Mach. Learn..

[34]  Huy L. Nguyen,et al.  Constrained Submodular Maximization: Beyond 1/e , 2016, 2016 IEEE 57th Annual Symposium on Foundations of Computer Science (FOCS).

[35]  Joseph Naor,et al.  A Unified Continuous Greedy Algorithm for Submodular Maximization , 2011, 2011 IEEE 52nd Annual Symposium on Foundations of Computer Science.

[36]  Ken-ichi Kawarabayashi,et al.  Optimal Budget Allocation: Theoretical Guarantee and Efficient Algorithm , 2014, ICML.