Improving Cluster-Based Algorithm Selection

There are a number of benefits to the ISAC methodology which have been touched upon in the previous chapters. In addition to those, it is also a transparent approach. Each cluster can be identified and analyzed for what makes it different from all the others. What are the features that are most different for a cluster that help differentiate it from its neighbors? Such analysis can then motivate understanding of why a particular solver performs well on one group of instances and not another. Identifying clusters where all solvers perform poorly can help direct research to developing novel solvers with a specific target benchmark. Clusters can also help identify what regions of the problem space have not yet been properly investigated. Yet with the introduction of new, more powerful techniques, this chapter focuses on presenting how the ISAC methodology can be improved without sacrificing its transparency. Specifically, the chapter will review some of the assumptions that have been made in the previous chapters and show whether they have been justified. It will then show how taking into account the performances of the solvers in a portfolio can help guide the clustering process.

[1]  Yoav Shoham,et al.  Towards a universal test suite for combinatorial auction algorithms , 2000, EC '00.

[2]  Carlos Ansótegui,et al.  A Gender-Based Genetic Algorithm for the Automatic Configuration of Algorithms , 2009, CP.

[3]  Thorsten Koch,et al.  Branching rules revisited , 2005, Oper. Res. Lett..

[4]  Yuri Malitsky,et al.  Algorithm Portfolios Based on Cost-Sensitive Hierarchical Clustering , 2013, IJCAI.

[5]  Kevin Leyton-Brown,et al.  Hydra: Automatically Configuring Algorithms for Portfolio-Based Selection , 2010, AAAI.

[6]  Alper Atamtürk,et al.  A study of the lot-sizing polytope , 2004, Math. Program..

[7]  Bart Selman,et al.  Domain-Independent Extensions to GSAT : Solving Large StructuredSatis ability , 1993 .

[8]  Alper Atamtürk,et al.  On the facets of the mixed–integer knapsack polyhedron , 2003, Math. Program..

[9]  Kevin Leyton-Brown,et al.  Performance Prediction and Automated Tuning of Randomized and Parametric Algorithms , 2006, CP.

[10]  Kerstin Eder,et al.  International Symposium on Code Generation and Optimization. CGO 2003 , 2003, International Symposium on Code Generation and Optimization, 2003. CGO 2003..

[11]  Nysret Musliu,et al.  Local Search Algorithm for Unicost Set Covering Problem , 2006, IEA/AIE.

[12]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[13]  Carlos Ansótegui,et al.  Disco - Novo - GoGo: Integrating Local Search and Complete Search with Restarts , 2006, AAAI.

[14]  William M. Rand,et al.  Objective Criteria for the Evaluation of Clustering Methods , 1971 .

[15]  Ashiqur R. KhudaBukhsh,et al.  SATenstein: automatically building local search SAT solvers from components , 2009, IJCAI 2009.

[16]  David H. Wolpert,et al.  No free lunch theorems for optimization , 1997, IEEE Trans. Evol. Comput..

[17]  Bart Selman,et al.  Evidence for Invariants in Local Search , 1997, AAAI/IAAI.

[18]  Andrea Lodi,et al.  MIPLIB 2010 , 2011, Math. Program. Comput..

[19]  George C. Runger,et al.  Using Experimental Design to Find Effective Parameter Settings for Heuristics , 2001, J. Heuristics.

[20]  Susan L. Epstein,et al.  The Adaptive Constraint Engine , 2002, CP.

[21]  Michel Lemaître,et al.  Branch and Bound Algorithm Selection by Performance Prediction , 1998, AAAI/IAAI.

[22]  Yoav Shoham,et al.  Understanding Random SAT: Beyond the Clauses-to-Variables Ratio , 2004, CP.

[23]  Kevin P. Murphy,et al.  Time-Bounded Sequential Parameter Optimization , 2010, LION.

[24]  Barry O'Sullivan,et al.  Tuning Parameters of Large Neighborhood Search for the Machine Reassignment Problem , 2013, CPAIOR.

[25]  Andrew W. Moore,et al.  X-means: Extending K-means with Efficient Estimation of the Number of Clusters , 2000, ICML.

[26]  Charles S. ReVelle,et al.  The Location of Emergency Service Facilities , 1971, Oper. Res..

[27]  Steven Minton,et al.  Automatically configuring constraint satisfaction programs: A case study , 1996, Constraints.

[28]  Barry O'Sullivan,et al.  Comparing Solution Methods for the Machine Reassignment Problem , 2012, CP.

[29]  H. Abdi,et al.  Principal component analysis , 2010 .

[30]  S. P. Lloyd,et al.  Least squares quantization in PCM , 1982, IEEE Trans. Inf. Theory.

[31]  Alex S. Fukunaga,et al.  Automated Discovery of Local Search Heuristics for Satisfiability Testing , 2008, Evolutionary Computation.

[32]  Predrag Janicic,et al.  Instance-Based Selection of Policies for SAT Solvers , 2009, SAT.

[33]  G. Dantzig,et al.  THE DECOMPOSITION ALGORITHM FOR LINEAR PROGRAMS , 1961 .

[34]  Steven David Prestwich,et al.  Random Walk with Continuously Smoothed Variable Weights , 2005, SAT.

[35]  David Pearce,et al.  Equilibrium logic , 2006, Annals of Mathematics and Artificial Intelligence.

[36]  Stephen F. Smith,et al.  Combining Multiple Heuristics Online , 2007, AAAI.

[37]  Michail G. Lagoudakis,et al.  Learning to Select Branching Rules in the DPLL Procedure for Satisfiability , 2001, Electron. Notes Discret. Math..

[38]  Toby Walsh,et al.  Morphing: Combining Structure and Randomness , 1999, AAAI/IAAI.

[39]  Andrew W. Moore,et al.  Learning Evaluation Functions to Improve Optimization by Local Search , 2001, J. Mach. Learn. Res..

[40]  Bart Selman,et al.  Problem Structure in the Presence of Perturbations , 1997, AAAI/IAAI.

[41]  David G. Lowe,et al.  Fast Approximate Nearest Neighbors with Automatic Algorithm Configuration , 2009, VISAPP.

[42]  Kate Smith-Miles,et al.  Cross-disciplinary perspectives on meta-learning for algorithm selection , 2009, CSUR.

[43]  John Thornton,et al.  Additive versus Multiplicative Clause Weighting for SAT , 2004, AAAI.

[44]  Alper Atamtürk,et al.  Flow pack facets of the single node fixed-charge flow polytope , 2001, Oper. Res. Lett..

[45]  F. Hutter,et al.  ParamILS: an automatic algorithm configuration framework , 2009 .

[46]  Yuri Malitsky,et al.  Instance-Specific Algorithm Configuration as a Method for Non-Model-Based Portfolio Generation , 2012, CPAIOR.

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

[48]  Francis J. Vasko,et al.  Optimal Selection of Ingot Sizes Via Set Covering , 1987, Oper. Res..

[49]  M. Mézard,et al.  Survey propagation: An algorithm for satisfiability , 2005 .

[50]  M. Padberg,et al.  Solving airline crew scheduling problems by branch-and-cut , 1993 .

[51]  Yuri Malitsky,et al.  ISAC - Instance-Specific Algorithm Configuration , 2010, ECAI.

[52]  Efthymios Housos,et al.  Automatic Optimization of Subproblems in Scheduling Airline Crews , 1997 .

[53]  Marijn J. H. Heule,et al.  March_eq: Implementing Additional Reasoning into an Efficient Look-Ahead SAT Solver , 2004, SAT (Selected Papers.

[54]  Tad Hogg,et al.  An Economics Approach to Hard Computational Problems , 1997, Science.

[55]  Mihai Oltean,et al.  Evolving Evolutionary Algorithms Using Linear Genetic Programming , 2005, Evolutionary Computation.

[56]  Yoav Shoham,et al.  A portfolio approach to algorithm select , 2003, IJCAI 2003.

[57]  Greg Hamerly,et al.  Learning the k in k-means , 2003, NIPS.

[58]  Serdar Kadioglu,et al.  Dialectic Search , 2009, CP.

[59]  Stephen F. Smith,et al.  New Techniques for Algorithm Portfolio Design , 2008, UAI.

[60]  Ian H. Witten,et al.  The WEKA data mining software: an update , 2009, SKDD.

[61]  Henry A. Kautz,et al.  Auto-Walksat: A Self-Tuning Implementation of Walksat , 2001, Electron. Notes Discret. Math..

[62]  Meinolf Sellmann,et al.  The Accuracy of Search Heuristics: An Empirical Study on Knapsack Problems , 2008, CPAIOR.

[63]  Yuri Malitsky,et al.  Stochastic Offline Programming , 2009, ICTAI.

[64]  Holger H. Hoos,et al.  Scaling and Probabilistic Smoothing: Efficient Dynamic Local Search for SAT , 2002, CP.

[65]  H. Terashima-Marín,et al.  Evolution of Constraint Satisfaction strategies in examination timetabling , 1999 .

[66]  Osamu Watanabe,et al.  On a Positive Instance Generation for the 3-Satisfiability (3SAT) Problem , 1996 .

[67]  Matteo Fischetti,et al.  Algorithms for railway crew management , 1997, Math. Program..

[68]  Kevin Leyton-Brown,et al.  : The Design and Analysis of an Algorithm Portfolio for SAT , 2007, CP.

[69]  Donald W. Loveland,et al.  A machine program for theorem-proving , 2011, CACM.

[70]  Roberto Battiti,et al.  The Reactive Tabu Search , 1994, INFORMS J. Comput..

[71]  Frank Hutter,et al.  Parameter Adjustment Based on Performance Prediction: Towards an Instance-Aware Problem Solver , 2005 .

[72]  Philippe Refalo,et al.  Impact-Based Search Strategies for Constraint Programming , 2004, CP.

[73]  Marek Petrik,et al.  Learning Static Parallel Portfolios of Algorithms , 2006, ISAIM.

[74]  Leo Breiman,et al.  Bagging Predictors , 1996, Machine Learning.

[75]  Andrea Lodi,et al.  Mixed Integer Programming Library version 5 , 2011 .

[76]  Martin W. P. Savelsbergh,et al.  Valid inequalities for problems with additive variable upper bounds , 2001, Math. Program..

[77]  Manuel Laguna,et al.  Fine-Tuning of Algorithms Using Fractional Experimental Designs and Local Search , 2006, Oper. Res..

[78]  Bart Selman,et al.  Algorithm portfolios , 2001, Artif. Intell..

[79]  Holger H. Hoos,et al.  An adaptive noise mechanism for walkSAT , 2002, AAAI/IAAI.

[80]  CHARLES AUDET,et al.  Finding Optimal Algorithmic Parameters Using Derivative-Free Optimization , 2006, SIAM J. Optim..

[81]  Horst Samulowitz,et al.  Learning to Solve QBF , 2007, AAAI.

[82]  Thomas Stützle,et al.  A Racing Algorithm for Configuring Metaheuristics , 2002, GECCO.

[83]  David A. Padua,et al.  Optimizing sorting with genetic algorithms , 2005, International Symposium on Code Generation and Optimization.

[84]  Martin Gebser,et al.  Solution Enumeration for Projected Boolean Search Problems , 2009, CPAIOR.

[85]  Luca Pulina,et al.  Collaborative Expert Portfolio Management , 2010, AAAI.

[86]  Kevin Leyton-Brown,et al.  SATzilla2009: an Automatic Algorithm Portfolio for SAT , 2008 .

[87]  Risto Miikkulainen,et al.  Latent class models for algorithm portfolio methods , 2010, AAAI 2010.

[88]  Kaile Su,et al.  Configuration Checking with Aspiration in Local Search for SAT , 2012, AAAI.

[89]  Yoav Shoham,et al.  Boosting as a Metaphor for Algorithm Design , 2003, CP.

[90]  Jürgen Schmidhuber,et al.  Dynamic Algorithm Portfolios , 2006, AI&M.