CS-TSSOS: Correlative and term sparsity for large-scale polynomial optimization

This work proposes a new moment-SOS hierarchy, called CS-TSSOS, for solving large-scale sparse polynomial optimization problems. Its novelty is to exploit simultaneously correlative sparsity and term sparsity by combining advantages of two existing frameworks for sparse polynomial optimization. The former is due to Waki et al. while the latter was initially proposed by Wang et al. and later exploited in the TSSOS hierarchy. In doing so we obtain CS-TSSOS -- a two-level hierarchy of semidefinite programming relaxations with (i), the crucial property to involve quasi block-diagonal matrices and (ii), the guarantee of convergence to the global optimum. We demonstrate its efficiency on several large-scale instances of the celebrated Max-Cut problem and the important industrial optimal power flow problem, involving up to several thousands of variables and ten thousands of constraints.

[1]  J. Gallier.  Quadratic Optimization Problems , 2020, Linear Algebra and Optimization with Applications to Machine Learning.

[2]  Matteo Tacchi,et al.  Approximating regions of attraction of a sparse polynomial differential system , 2019, ArXiv.

[3]  Alan Edelman,et al.  Julia: A Fresh Approach to Numerical Computing , 2014, SIAM Rev..

[4]  Franz Rendl,et al.  Solving Max-Cut to optimality by intersecting semidefinite and polyhedral relaxations , 2009, Math. Program..

[5]  David Grimm,et al.  A note on the representation of positive polynomials with structured sparsity , 2006, math/0611498.

[6]  G. Stengle.  A nullstellensatz and a positivstellensatz in semialgebraic geometry , 1974 .

[7]  Gennadiy Averkov,et al.  Optimal Size of Linear Matrix Inequalities in Semidefinite Approaches to Polynomial Optimization , 2018, SIAM J. Appl. Algebra Geom..

[8]  L. Rodman,et al.  Positive semidefinite matrices with a given sparsity pattern , 1988 .

[9]  Jie Wang,et al.  Exploiting Sparsity in Complex Polynomial Optimization , 2021, J. Optim. Theory Appl..

[10]  R. Möhring.  Algorithmic graph theory and perfect graphs , 1986 .

[11]  Daniel K. Molzahn,et al.  Lasserre Hierarchy for Large Scale Polynomial Optimization in Real and Complex Variables , 2017, SIAM J. Optim..

[12]  J. Lasserre.  An Introduction to Polynomial and Semi-Algebraic Optimization , 2015 .

[13]  Masakazu Muramatsu,et al.  Sums of Squares and Semidefinite Programming Relaxations for Polynomial Optimization Problems with Structured Sparsity , 2004 .

[14]  B. Peyton,et al.  An Introduction to Chordal Graphs and Clique Trees , 1993 .

[15]  Jean B. Lasserre,et al.  Global Optimization with Polynomials and the Problem of Moments , 2000, SIAM J. Optim..

[16]  J. Lasserre,et al.  A Sparse Version of Reznick's Positivstellensatz , 2020, Math. Oper. Res..

[17]  Giovanni Rinaldi,et al.  SpeeDP: an algorithm to compute SDP bounds for very large Max-Cut instances , 2012, Mathematical Programming.

[18]  Masakazu Muramatsu,et al.  SparsePOP: a Sparse Semidefinite Programming Relaxation of Polynomial Optimization Problems , 2005 .

[19]  Monique Laurent,et al.  A Comparison of the Sherali-Adams, Lovász-Schrijver, and Lasserre Relaxations for 0-1 Programming , 2003, Math. Oper. Res..

[20]  Paul Rolland,et al.  Lipschitz constant estimation of Neural Networks via sparse polynomial optimization , 2020, ICLR.

[21]  Timo de Wolff,et al.  Amoebas, nonnegative polynomials and sums of squares supported on circuits , 2014, 1402.0462.

[22]  Igor Klep,et al.  Sparse noncommutative polynomial optimization , 2019 .

[23]  J. Krivine,et al.  Anneaux préordonnés , 1964 .

[24]  Antonis Papachristodoulou,et al.  Decomposed Structured Subsets for Semidefinite and Sum-of-Squares Optimization , 2019, ArXiv.

[25]  George A. Constantinides,et al.  Certified Roundoff Error Bounds Using Semidefinite Programming , 2015, ACM Trans. Math. Softw..

[26]  Knud D. Andersen,et al.  The Mosek Interior Point Optimizer for Linear Programming: An Implementation of the Homogeneous Algorithm , 2000 .

[27]  Volkan Cevher,et al.  Scalable Semidefinite Programming , 2019, SIAM J. Math. Data Sci..

[28]  B. Reznick.  Extremal PSD forms with few terms , 1978 .

[29]  Victor Magron,et al.  A second order cone characterization for sums of nonnegative circuits , 2020, ISSAC.

[30]  J. Lasserre,et al.  Detecting global optimality and extracting solutions in GloptiPoly , 2003 .

[31]  Victor Magron,et al.  Interval Enclosures of Upper Bounds of Roundoff Errors Using Semidefinite Programming , 2016, ACM Trans. Math. Softw..

[32]  Joachim Dahl,et al.  On the robustness and scalability of semidefinite relaxation for optimal power flow problems , 2018, Optimization and Engineering.

[33]  Charles R. Johnson,et al.  Positive definite completions of partial Hermitian matrices , 1984 .

[34]  Victor Magron,et al.  Chordal-TSSOS: A Moment-SOS Hierarchy That Exploits Term Sparsity with Chordal Extension , 2020, SIAM J. Optim..

[35]  Bican Xia,et al.  A New Sparse SOS Decomposition Algorithm Based on Term Sparsity , 2018, ISSAC.

[36]  Jie Wang,et al.  Exploiting term sparsity in noncommutative polynomial optimization , 2020, Computational Optimization and Applications.

[37]  Parikshit Shah,et al.  Relative Entropy Relaxations for Signomial Optimization , 2014, SIAM J. Optim..

[38]  M. Ferris,et al.  The Power Grid Library for Benchmarking AC Optimal Power Flow Algorithms , 2019, ArXiv.

[39]  Jiawang Nie,et al.  Optimality conditions and finite convergence of Lasserre’s hierarchy , 2012, Math. Program..

[40]  Martin S. Andersen,et al.  Chordal Graphs and Semidefinite Optimization , 2015, Found. Trends Optim..

[41]  Jean B. Lasserre,et al.  Convergent SDP-Relaxations for Polynomial Optimization with Sparsity , 2006, ICMS.

[42]  R. Saigal,et al.  Handbook of semidefinite programming : theory, algorithms, and applications , 2000 .

[43]  Blair J R S,et al.  Introduction to Chordal Graphs and Clique Trees, in Graph Theory and Sparse Matrix Computation , 1997 .

[44]  Amir Ali Ahmadi,et al.  DSOS and SDSOS optimization: LP and SOCP-based alternatives to sum of squares optimization , 2014, 2014 48th Annual Conference on Information Sciences and Systems (CISS).

[45]  D. Henrion,et al.  Exploiting Sparsity for Semi-Algebraic Set Volume Computation , 2019, Foundations of Computational Mathematics.

[46]  Jie Wang,et al.  TSSOS: a Julia library to exploit sparsity for large-scale polynomial optimization , 2021, ArXiv.

[47]  Jie Wang,et al.  TSSOS: A Moment-SOS hierarchy that exploits term sparsity , 2020 .

[48]  Victor Magron,et al.  Polynomial Optimization for Bounding Lipschitz Constants of Deep Networks , 2020, ArXiv.

[49]  Jean Maeght,et al.  Improving Clique Decompositions of Semidefinite Relaxations for Optimal Power Flow Problems , 2019, ArXiv.

[50]  Jean B. Lasserre,et al.  A bounded degree SOS hierarchy for polynomial optimization , 2015, EURO J. Comput. Optim..

[51]  D. R. Fulkerson,et al.  Incidence matrices and interval graphs , 1965 .

[52]  Thorsten Theobald,et al.  Exploiting Symmetries in SDP-Relaxations for Polynomial Optimization , 2011, Math. Oper. Res..

[53]  Amir Ali Ahmadi,et al.  DSOS and SDSOS Optimization: More Tractable Alternatives to Sum of Squares and Semidefinite Optimization , 2017, SIAM J. Appl. Algebra Geom..

[54]  Johan Löfberg,et al.  Pre- and Post-Processing Sum-of-Squares Programs in Practice , 2009, IEEE Transactions on Automatic Control.