Empirical Hardness Models for Combinatorial Auctions

Shoham In this chapter we consider the empirical hardness of the winner determination problem. We identify distribution-nonspecific features of data instances and then use statistical regression techniques to learn, evaluate and interpret a function from these features to the predicted hardness of an instance, focusing mostly on ILOG's CPLEX solver. We also describe two applications of these models: building an algorithm portfolio that selects among different WDP algorithms, and inducing test distributions that are harder for this algorithm portfolio.

[1]  David Maxwell Chickering,et al.  A Bayesian Approach to Tackling Hard Computational Problems (Preliminary Report) , 2001, Electron. Notes Discret. Math..

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

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

[4]  Eric R. Ziegel,et al.  The Elements of Statistical Learning , 2003, Technometrics.

[5]  Peter C. Cheeseman,et al.  Where the Really Hard Problems Are , 1991, IJCAI.

[6]  Tuomas Sandholm,et al.  Algorithm for optimal winner determination in combinatorial auctions , 2002, Artif. Intell..

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

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

[9]  Richard E. Korf,et al.  Complexity Analysis of Admissible Heuristic Search , 1998, AAAI/IAAI.

[10]  Weixiong Zhang.  State-space search - algorithms, complexity, extensions, and applications , 1999 .

[11]  Toby Walsh,et al.  Backbones in Optimization and Approximation , 2001, IJCAI.

[12]  Hector J. Levesque,et al.  Generating Hard Satisfiability Problems , 1996, Artif. Intell..

[13]  Yoav Shoham,et al.  Learning the Empirical Hardness of Optimization Problems: The Case of Combinatorial Auctions , 2002, CP.

[14]  Daniel Lehmann,et al.  Optimal solutions for multi-unit combinatorial auctions: branch and bound heuristics , 2000, EC '00.

[15]  Daniel Lehmann,et al.  Linear Programming helps solving large multi-unit combinatorial auctions , 2002, ArXiv.

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

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

[18]  John R. Rice,et al.  The Algorithm Selection Problem , 1976, Adv. Comput..

[19]  Eric Horvitz,et al.  Restart Policies with Dependence among Runs: A Dynamic Programming Approach , 2002, CP.

[20]  Rémi Monasson,et al.  Determining computational complexity from characteristic ‘phase transitions’ , 1999, Nature.

[21]  Ivana Kruijff-Korbayová,et al.  A Portfolio Approach to Algorithm Selection , 2003, IJCAI.

[22]  Bart Selman,et al.  Generating Satisfiable Problem Instances , 2000, AAAI/IAAI.

[23]  Timothy J. Robinson,et al.  Sequential Monte Carlo Methods in Practice , 2003 .

[24]  Yoav Shoham,et al.  Taming the Computational Complexity of Combinatorial Auctions: Optimal and Approximate Approaches , 1999, IJCAI.

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

[26]  Michail G. Lagoudakis,et al.  Algorithm Selection using Reinforcement Learning , 2000, ICML.