1 Combinatorial Problems in Cooperative Control: Complexity and Scalability Carla Gomes and Bart Selman Cornell University Muri Meeting March 2002.

Slides:



Advertisements
Similar presentations
1 Integration of Artificial Intelligence and Operations Research Techniques for Combinatorial Problems Carla P. Gomes Cornell University
Advertisements

1 Chapter 11 Here we see cases where the basic LP model can not be used.
© Imperial College London Eplex: Harnessing Mathematical Programming Solvers for Constraint Logic Programming Kish Shen and Joachim Schimpf IC-Parc.
Probabilistic Planning (goal-oriented) Action Probabilistic Outcome Time 1 Time 2 Goal State 1 Action State Maximize Goal Achievement Dead End A1A2 I A1.
1 Backdoor Sets in SAT Instances Ryan Williams Carnegie Mellon University Joint work in IJCAI03 with: Carla Gomes and Bart Selman Cornell University.
EE 553 Integer Programming
Dynamic Restarts Optimal Randomized Restart Policies with Observation Henry Kautz, Eric Horvitz, Yongshao Ruan, Carla Gomes and Bart Selman.
Lecture 10: Integer Programming & Branch-and-Bound
© 2011 IBM Corporation 1 Guiding Combinatorial Optimization with UCT Ashish Sabharwal and Horst Samulowitz IBM Watson Research Center (presented by Raghuram.
Progress in Linear Programming Based Branch-and-Bound Algorithms
Computational problems, algorithms, runtime, hardness
CPSC 322, Lecture 15Slide 1 Stochastic Local Search Computer Science cpsc322, Lecture 15 (Textbook Chpt 4.8) February, 6, 2009.
Integer Programming and Branch and Bound Brian C. Williams November 15 th, 17 th, 2004 Adapted from slides by Eric Feron, , 2002.
Statistical Regimes Across Constrainedness Regions Carla P. Gomes, Cesar Fernandez Bart Selman, and Christian Bessiere Cornell University Universitat de.
CP Formal Models of Heavy-Tailed Behavior in Combinatorial Search Hubie Chen, Carla P. Gomes, and Bart Selman
1 Optimisation Although Constraint Logic Programming is somehow focussed in constraint satisfaction (closer to a “logical” view), constraint optimisation.
Heavy-Tailed Behavior and Search Algorithms for SAT Tang Yi Based on [1][2][3]
CPSC 322, Lecture 12Slide 1 CSPs: Search and Arc Consistency Computer Science cpsc322, Lecture 12 (Textbook Chpt ) January, 29, 2010.
Computational Methods for Management and Economics Carla Gomes
Connections in Networks: A Hybrid Approach Carla P. Gomes, Willem-Jan van Hoeve, Ashish Sabharwal Cornell University CP-AI-OR Conference, May 2008 Paris,
Formal Complexity Analysis of Mobile Problems & Communication and Computation in Distributed Sensor Networks in Distributed Sensor Networks Carla P. Gomes.
1 Backdoors To Typical Case Complexity Ryan Williams Carnegie Mellon University Joint work with: Carla Gomes and Bart Selman Cornell University.
Carla P. Gomes CS4700 CS 4700: Foundations of Artificial Intelligence Carla P. Gomes Module: Randomization in Complete Tree Search.
Integer Programming Difference from linear programming –Variables x i must take on integral values, not real values Lots of interesting problems can be.
CP-AI-OR-02 Gomes & Shmoys 1 The Promise of LP to Boost CSP Techniques for Combinatorial Problems Carla P. Gomes David Shmoys
1 BLACKBOX: A New Approach to the Application of Theorem Proving to Problem Solving Bart Selman Cornell University Joint work with Henry Kautz AT&T Labs.
Distributed Combinatorial Optimization
1 Planning and Scheduling to Minimize Tardiness John Hooker Carnegie Mellon University September 2005.
Carla P. Gomes School on Optimization CPAIOR02 Exploiting Structure and Randomization in Combinatorial Search Carla P. Gomes
Lukas Kroc, Ashish Sabharwal, Bart Selman Cornell University, USA SAT 2010 Conference Edinburgh, July 2010 An Empirical Study of Optimal Noise and Runtime.
Multi-vehicle Cooperative Control Raffaello D’Andrea Mechanical & Aerospace Engineering Cornell University u Progress on RoboFlag Test-bed u MLD approach.
LP formulation of Economic Dispatch
Daniel Kroening and Ofer Strichman Decision Procedures An Algorithmic Point of View Deciding ILPs with Branch & Bound ILP References: ‘Integer Programming’
(Not in text).  An LP with additional constraints requiring that all the variables be integers is called an all-integer linear program (IP).  The LP.
Decision Procedures An Algorithmic Point of View
Distributions of Randomized Backtrack Search Key Properties: I Erratic behavior of mean II Distributions have “heavy tails”.
1 Outline:  Outline of the algorithm  MILP formulation  Experimental Results  Conclusions and Remarks Advances in solving scheduling problems with.
Types of IP Models All-integer linear programs Mixed integer linear programs (MILP) Binary integer linear programs, mixed or all integer: some or all of.
CPS 270: Artificial Intelligence More search: When the path to the solution doesn’t matter Instructor: Vincent.
MILP algorithms: branch-and-bound and branch-and-cut
Tobias Achterberg Konrad-Zuse-Zentrum für Informationstechnik Berlin Branching SCIP Workshop at ZIB October 2007.
Heavy-Tailed Phenomena in Satisfiability and Constraint Satisfaction Problems by Carla P. Gomes, Bart Selman, Nuno Crato and henry Kautz Presented by Yunho.
CSE 589 Part VI. Reading Skiena, Sections 5.5 and 6.8 CLR, chapter 37.
15.053Tuesday, April 9 Branch and Bound Handouts: Lecture Notes.
1 Outline:  Optimization of Timed Systems  TA-Modeling of Scheduling Tasks  Transformation of TA into Mixed-Integer Programs  Tree Search for TA using.
Conformant Probabilistic Planning via CSPs ICAPS-2003 Nathanael Hyafil & Fahiem Bacchus University of Toronto.
1 Iterative Integer Programming Formulation for Robust Resource Allocation in Dynamic Real-Time Systems Sethavidh Gertphol and Viktor K. Prasanna University.
Divide and Conquer Optimization problem: z = max{cx : x  S}
Quality of LP-based Approximations for Highly Combinatorial Problems Lucian Leahu and Carla Gomes Computer Science Department Cornell University.
Integer LP In-class Prob
Implicit Hitting Set Problems Richard M. Karp Erick Moreno Centeno DIMACS 20 th Anniversary.
1 Combinatorial Problems in Cooperative Control: Complexity and Scalability Carla P. Gomes and Bart Selman Cornell University Muri Meeting June 2002.
September 28, 2000 Improved Simultaneous Data Reconciliation, Bias Detection and Identification Using Mixed Integer Optimization Methods Presented by:
Introduction to Integer Programming Integer programming models Thursday, April 4 Handouts: Lecture Notes.
Constraint Programming for the Diameter Constrained Minimum Spanning Tree Problem Thiago F. Noronha Celso C. Ribeiro Andréa C. Santos.
Formal Complexity Analysis of RoboFlag Drill & Communication and Computation in Distributed Negotiation Algorithms in Distributed Negotiation Algorithms.
Biointelligence Lab School of Computer Sci. & Eng. Seoul National University Artificial Intelligence Chapter 8 Uninformed Search.
Decision Support Systems
Data Driven Resource Allocation for Distributed Learning
C. Kiekintveld et al., AAMAS 2009 (USC) Presented by Neal Gupta
C.-S. Shieh, EC, KUAS, Taiwan
Instructor: Vincent Conitzer
Introduction to Operations Research
MILP algorithms: branch-and-bound and branch-and-cut
MIP Tools Branch and Cut with Callbacks Lazy Constraint Callback
Search.
Search.
Solution methods for NP-hard Discrete Optimization Problems
Area Coverage Problem Optimization by (local) Search
Discrete Optimization
Presentation transcript:

1 Combinatorial Problems in Cooperative Control: Complexity and Scalability Carla Gomes and Bart Selman Cornell University Muri Meeting March 2002

2 We are investigating how to scale up solutions of the ROBOFLAG Drill focusing on: - Mixed Integer Program (MIP) formulations - Randomization - Approximation methods - Portfolios of Algorithms - Combining MIP and constraint search techniques.

3 Problem Representation ROBOFLAG Drill Formulation by Raff D’Andrea and Matt Earl. Problem is hybrid, combining discrete and continuous components, with multiple constraints. Represented as a mixed logical system (MLD) in which the objective is to compute optimal control policies that minimize the total score of the game. Mathematical Formulation of the Optimization Problem  Mixed Integer Linear Program

4 Scaling Up Mixed Integer Linear Program Formulations (MILP) Standard approach for solving MILP: Branch and Bound How can we improve upon Branch and Bound strategies? Ideas: Randomization Different search strategies for node selection Portfolios of algorithms

5 Branch & Bound: Depth First vs. Best bound Critical to performance of Branch & Bound is the way in which the next node to be expanded is selected. Standard approach: Best-bound --- select the node with the best LP bound Alternative: Depth-first --- often quickly reaches an integer solution (may take longer to produce an overall optimal value) Tradeoffs between these choices depend on underlying problem stucture (Gomes et al. 2001).

6 ROBOFLAG Testbed Depth First search works well. Problems that could not be solved before with best bound using were solved with depth first. Current largest problem solved with CPLEX using Depth First Search (8 attackers and 3 defenders): Integer variables = 4040 Continuous variables 400 Constraints constraints Time secs (Matt Earl 2002)

7 Much room for improvement… We are not yet incorporating any randomization or discrete constraint propagation techniques. Nor are we yet exploiting parallelism using a portfolio approach. Doing so should allow us to solve problems at least one or two orders of magnitude larger. (100,000 to 500,000 vars and 1,000,000+ constraints) Also, we should be able to include more complex constraints.

8 Other Formulations for Solving the Control Optimization Problem Encodings that provide “tighter” relaxations for the LP problem. Approximate representations using abstractions (“synthesize larger movements / trajecturies”). Less compact representations may allow for more propagation and scale up better. Constraint Satisfaction Problem (CSP) formulations. (*) Hybrid CSP/LP formulation. Approximations based on LP randomized rounding. (*)Sat – the satisfiability problem is a particular case of CSP; however, we believe that SAT encodings may not scale up well in this domain.

9 Overall the Roboflag control problem provides an excellent test bed for the development of scalable techniques for complex optimization.

10 Auxiliary Slides Background on improvements on branch and bound using randomization and parallel portfolios.

11 Branch & Bound (Randomized) Solve linear relaxation of MIP Branch on the integer variables for which the solution of the LP relaxation is non-integer: apply a good heuristic (e.g., max infeasibility) for variable selection ( + randomization ) and create two new nodes (floor and ceiling of the fractional value) Once we have found an integer solution, its objective value can be used to prune other nodes, whose relaxations have worse values

12 The performance of randomized Branch and Bound varies dramatically, on the same instance. In fact, the run time distributions often exhibit long tails (Heavy-tailed Distributions)

13 Heavy-tailed behavior of Depth-first

14 So, how can we take advantage of the high variability of randomized methods? - restart strategies - portfolio strategies

15 Algorithm Portfolio Design

16 Motivation The runtime and performance of randomized algorithms can vary dramatically on the same instance and on different instances. Goal: Improve the performance of different algorithms by combining them into a portfolio to exploit their relative strengths.

17 Portfolio of Algorithms A portfolio of algorithm is a collection of algorithms and / or copies of the same algorithm running interleaved or on different processors. Goal: to improve on the performance of the component algorithms in terms of: expected computational cost “risk” (variance) Efficient Set or Efficient Frontier: set of portfolios that are best in terms of expected value and risk.

18 Depth-first vs. Best-bound (logistics planning) Number of nodes Cumulative Frequencies Depth-First ~50% Best-Bound ~30%

19 Depth-First and Best and Bound do not dominate each other overall. What if we have more than one processors or if we interleave processes on a single processor?

20 Portfolio for heavy-tailed search procedures (2 processors) 0 DF / 2 BB 2 DF / 0 BB Standard deviation of run time of portfolios Expected run time of portfolios

21 Portfolio for heavy-tailed search procedures (20 processors) 0 DF / 20 BB 20 DF / 0 BB Standard deviation of run time of portfolios Expected run time of portfolios The optimal strategy is to run Depth First on the 20 processors!

22 Optimal collective behavior can emerge from suboptimal individual behavior.

23 A portfolio approach can lead to substantial improvements in the expected cost and risk of stochastic algorithms, especially in the presence of heavy-tailed phenomena.