The Seesaw Algorithm: Function Optimization Using Implicit Hitting Sets

The paper introduces the Seesaw algorithm, which explores the Pareto frontier of two given functions. The algorithm is complete and generalizes the well-known implicit hitting set paradigm. The first given function determines a cost of a hitting set and is optimized by an exact solver. The second, called the oracle function, is treated as a black-box. This approach is particularly useful in the optimization of functions that are impossible to encode into an exact solver. We show the effectiveness of the algorithm in the context of static solver portfolio selection. The existing implicit hitting set paradigm is applied to cost function and an oracle predicate. Hence, the Seesaw algorithm generalizes this by enabling the oracle to be a function. The paper identifies two independent preconditions that guarantee the correctness of the algorithm. This opens a number of avenues for future research into the possible instantiations of the algorithm, depending on the cost and oracle functions used. 2012 ACM Subject Classification Computing methodologies → Optimization algorithms

[1]  Mikolás Janota,et al.  Abstraction-Based Algorithm for 2QBF , 2011, SAT.

[2]  Mikolás Janota,et al.  Quantified maximum satisfiability , 2015, Constraints.

[3]  Josef Urban,et al.  Hammering Mizar by Learning Clause Guidance , 2019, ArXiv.

[4]  Josef Urban,et al.  BliStrTune: hierarchical invention of theorem proving strategies , 2017, CPP.

[5]  Stasys Jukna.  Exponential lower bounds for semantic resolution , 1996, Proof Complexity and Feasible Arithmetics.

[6]  Mikolás Janota,et al.  Quantified Maximum Satisfiability: - A Core-Guided Approach , 2013, SAT.

[7]  Matti Järvisalo,et al.  Implicit Hitting Set Algorithms for Reasoning Beyond NP , 2016, KR.

[8]  Richard M. Karp,et al.  The Implicit Hitting Set Approach to Solve Combinatorial Optimization Problems with an Application to Multigenome Alignment , 2013, Oper. Res..

[9]  Joao Marques-Silva,et al.  Propositional Abduction with Implicit Hitting Sets , 2016, ECAI.

[10]  C. Hwang.  Multiple Objective Decision Making - Methods and Applications: A State-of-the-Art Survey , 1979 .

[11]  Josef Urban,et al.  ENIGMA Anonymous: Symbol-Independent Inference Guiding Machine (System Description) , 2020, IJCAR.

[12]  Mikolás Janota,et al.  On the query complexity of selecting minimal sets for monotone predicates , 2016, Artif. Intell..

[13]  Raymond Reiter,et al.  A Theory of Diagnosis from First Principles , 1986, Artif. Intell..

[14]  Josef Urban,et al.  MPTP 0.2: Design, Implementation, and Initial Experiments , 2006, Journal of Automated Reasoning.

[15]  Joao Marques-Silva,et al.  Horn Maximum Satisfiability: Reductions, Algorithms and Applications , 2017, EPIA.

[16]  Vasco M. Manquinho,et al.  Introducing Pareto Minimal Correction Subsets , 2017, SAT.

[17]  J. Hooker,et al.  Logic-based Benders decomposition , 2003 .

[18]  Jennifer Ryan,et al.  Finding the minimum weight IIS cover of an infeasible system of linear inequalities , 1996, Annals of Mathematics and Artificial Intelligence.

[19]  Jacques F. Benders,et al.  Partitioning procedures for solving mixed-variables programming problems , 2005, Comput. Manag. Sci..

[20]  Inês Lynce,et al.  Conflict-Driven Clause Learning SAT Solvers , 2009, Handbook of Satisfiability.

[21]  Mikolás Janota,et al.  Solving QBF with Counterexample Guided Refinement , 2012, SAT.

[22]  Joao Marques-Silva,et al.  Smallest MUS Extraction with Minimal Hitting Set Dualization , 2015, CP.

[23]  Josep Argelich,et al.  Boolean lexicographic optimization: algorithms & applications , 2011, Annals of Mathematics and Artificial Intelligence.

[24]  David S. Johnson,et al.  Computers and Intractability: A Guide to the Theory of NP-Completeness , 1978 .

[25]  Carsten Sinz,et al.  Overview and analysis of the SAT Challenge 2012 solver competition , 2015, Artif. Intell..

[26]  Fahiem Bacchus,et al.  Solving MAXSAT by Solving a Sequence of Simpler SAT Instances , 2011, CP.

[27]  Armin Biere,et al.  Implicit Hitting Set Algorithms for Maximum Satisfiability Modulo Theories , 2018, IJCAR.

[28]  Fahiem Bacchus,et al.  Exploiting the Power of mip Solvers in maxsat , 2013, SAT.

[29]  Kevin Leyton-Brown,et al.  Selection and Configuration of Parallel Portfolios , 2018, Handbook of Parallel Constraint Reasoning.

[30]  Alexander Nadel,et al.  On Optimizing a Generic Function in SAT , 2020, 2020 Formal Methods in Computer Aided Design (FMCAD).

[31]  Fahiem Bacchus,et al.  Solving Weighted CSPs by Successive Relaxations , 2013, CP.

[32]  Meir Kalech,et al.  Exploring the Duality in Conflict-Directed Model-Based Diagnosis , 2012, AAAI.

[33]  Eoin O'Mahony,et al.  Using Case-based Reasoning in an Algorithm Portfolio for Constraint Solving ? , 2008 .

[34]  Stephan Schulz,et al.  E - a brainiac theorem prover , 2002, AI Commun..

[35]  Jakob Nordström,et al.  A Cardinal Improvement to Pseudo-Boolean Solving , 2020, AAAI.

[36]  James A. Reggia,et al.  Diagnostic Expert Systems Based on a Set Covering Model , 1983, Int. J. Man Mach. Stud..

[37]  Christodoulos A. Floudas.  Generalized Benders Decomposition , 2009, Encyclopedia of Optimization.

[38]  Kevin Leyton-Brown,et al.  SATzilla: Portfolio-based Algorithm Selection for SAT , 2008, J. Artif. Intell. Res..

[39]  E. L. Ulungu,et al.  Multi‐objective combinatorial optimization problems: A survey , 1994 .