## Increasing the Efficiency of Simulated Annealing Search by Learning to Recognize (Un)Promising Runs (1994)

### Cached

### Download Links

- [www.normsadeh.com]
- [www.ri.cmu.edu]
- [www.normsadeh.com]
- DBLP

### Other Repositories/Bibliography

Venue: | Proceedings of the 12 th National Conference on Artificial Intelligence (AAAI'94 |

Citations: | 8 - 1 self |

### BibTeX

@INPROCEEDINGS{Nakakuki94increasingthe,

author = {Yoichiro Nakakuki and Norman Sadeh},

title = {Increasing the Efficiency of Simulated Annealing Search by Learning to Recognize (Un)Promising Runs},

booktitle = {Proceedings of the 12 th National Conference on Artificial Intelligence (AAAI'94},

year = {1994}

}

### Years of Citing Articles

### OpenURL

### Abstract

Simulated Annealing (SA) procedures can potentially yield near-optimal solutions to many difficult combinatorial optimization problems, though often at the expense of intensive computational efforts. The single most significant source of inefficiency in SA search is its inherent stochasticity, typically requiring that the procedure be rerun a large number of times before a near-optimal solution is found. This paper describes a mechanism that attempts to learn the structure of the search space over multiple SA runs on a given problem. Specifically, probability distributions are dynamically updated over multiple runs to estimate at different checkpoints how promising a SA run appears to be. Based on this mechanism, two types of criteria are developed that aim at increasing search efficiency: (1) a cuto $ criterion used to determine when to abandon unpromising runs and (2) restart criteria used to determine whether to start a fresh SA run or restart search in the middle of an earher run. Experimental results obtained on a class of complex job shop scheduling problems show (1) that SA can produce high quality solutions for this class of problems, if run a large number of times, and (2) that our learning mechanism can significantly reduce the computation time required to find high quality solutions to these problems. The results further indicate that, the closer one wants to be to the optimum, the larger the speedups. 1

### Citations

10964 | Computers and Intractability: A Guide to the Theory of NP-Completeness - Garey, Johnson - 1979 |

512 |
Optimization by Simulated Annealing: An Experimental Evaluation
- Johnson, Aragon, et al.
- 1991
(Show Context)
Citation Context ... SA procedures have been successfully applied to a variety of combinatorial optimiz.ation problems, including Traveling Salesman Problems [I]! Graph Partitioning Problems 161, Graph Coloring Problems =-=[7]-=-! Vehicle Routing Problems [14], Design of Integrat’cd Circuits, Minimum Makespan Scheduling Problems [9, 13, 191 as well as othcr C.OIIIplex scheduling problems (23]? often producing near-optimal sol... |

317 | Thermodynamical approach to the traveling salesman problem: An efficient simulation algorithm - Černý - 1985 |

168 | Learning with Delayed Rewards - Watkins - 1989 |

148 | Computers and Intractability: A guide to the theory of NP-Completeness - Gary, Johnson - 1979 |

114 | Look-ahead techniques for micro-opportunistic job shop scheduling
- Sadeh
- 1996
(Show Context)
Citation Context ...regnlar objecdives sucb as Minimum Makespan. It can he shown that the neighborhoods used in these earlier studies are no1 adequat,e to deal n7it.h irregular objectives such as the one considered here =-=[IS]-=-. 4An operation is said to be ”right(1eft)-shiftable” if its start time can be increrrsed (decreased) by one time unit without ovcrlapping with another operation. 51n our implementation, exchanging tw... |

112 |
Meta-Strategy Simulated Annealing and Tabu Search Algorithms for the Vehicle Routing Problem
- Osman
- 1992
(Show Context)
Citation Context ...ssfully applied to a variety of combinatorial optimiz.ation problems, including Traveling Salesman Problems [I]! Graph Partitioning Problems 161, Graph Coloring Problems [7]! Vehicle Routing Problems =-=[14]-=-, Design of Integrat’cd Circuits, Minimum Makespan Scheduling Problems [9, 13, 191 as well as othcr C.OIIIplex scheduling problems (23]? often producing near-optimal solut,ions, t,hough at the expense... |

100 | Rescheduling with Iterative Repair - Zweben, Davis, et al. - 1992 |

84 | Heuristic Scheduling Systems - Morton, Pentico - 1993 |

66 | A Review of Production Scheduling - Graves - 1981 |

55 | Micro-Opportunistic Scheduling: The MicroBoss Factory scheduler” Ch - Sadeh - 1994 |

43 | Searching With Probabilities - Palay - 1521 |

37 | A Controlled Search Simulated Annealing Method for the General Jobshop Scheduling Problem - Matsuo, Suh, et al. - 1988 |

37 | Simulated Annealing for Permutation Flow-Shop Scheduling,” OMEGA Int - Osman, Potts - 1989 |

33 | Priority rules for job shops with weighted tardiness costs - Vopsalainon, Morton - 1987 |

11 | SCHEDSTAR: a price-based shop scheduling module - Morton, Lawrence, et al. - 1988 |

10 | Focused Simulated Annealing Search: An Application to Job Shop Scheduling - Sadeh, Nakakuki - 1994 |

10 | Adaptive Learning of Decision-Theoretic Search - Wefald, Russel - 1989 |

3 | Lenstra “Job Shop Scheduling by Simulated Annealing - Laarhoven, Aarts, et al. - 1992 |

2 | Heuristic Scheduling Systems,” Wiley Series in Engineering and Technology Management - Morton, Pentico - 1993 |

1 |
Thermodynamical Approach lo the Traveling Salcsman Problem: An Efficient Siniulation Algorithm
- Cerny
- 1985
(Show Context)
Citation Context ...ch procedure t,liat generalizes iterat,ive improvement approaches to Combinatorial opt,imization by s0metime.s accepting transitions to lower quality solutions to avoid getting trapped in local minima=-=[8, I]-=-. SA procedures have been successfully applied to a variety of combinatorial optimiz.ation problems, including Traveling Salesman Problems [I]! Graph Partitioning Problems 161, Graph Coloring Problems... |

1 |
Tabu Search,” To appear as a Chapter in Modern Heuristic Techniques for Combinatorial Problems
- Glover, Laguna
- 1992
(Show Context)
Citation Context ...of ”Tabu” mechanisms that attempt to increase the efficiency of SA and other neighborhood search procedures by maintaining a selective history of search states encountered earlier during the same Tun =-=[4]-=-. This history is then used t,o dynamically derive “t,abu restrictions” or ”aspirations”, that guide search, prevent,ing it, for instance, from revisit,ing areas of the search space it just explored. ... |

1 | Increasing the Efficiency of Simulated Plnncaling Search by Learning to Recognize (Un)Promising Runs - Nakakuki, SadehJJ - 1994 |

1 | Simulated Annealing for Permutation Flow-Shop Schcduling,’ OMEGA ht - OsmanJII, Pot, et al. - 1989 |

1 | Rescheduling with hrative Repair - Zwebcn, Davis, et al. - 1992 |