Broderick Crawford

Adaptive enumeration strategies and metabacktracks for constraint solving

In Constraint Programming, enumeration strategies are crucial for resolution performances. The effect of strategies is generally unpredictable. In a previous work, we proposed to dynamically change strategies showing bad performances, and to use metabacktrack to restore better states when bad decisions were made. In this paper, we design and evaluate strategies to improve resolution performances of a set of problems. Experimental results show the effectiveness of our approach.

Using Constraint Programming to solve Sudoku Puzzles

In constraint programming, enumeration strategies are crucial for resolution performances. In this work, we model the known NP-complete problems Latin Square, Magic Square and Sudoku as a constraint satisfaction problems. We solve them with constraint programming comparing the performance of different variable and value selection heuristics in its enumeration phase.

Hypercube FrameWork for ACO applied to timetabling

We present a resolution technique of the University course Timetabling problem (UCTP), this technique is based in the implementation of Hypercube framework using the Max-Min Ant System. We presented the structure of the problem and the design of resolution using this framework.

A New ACO Transition Rule for Set Partitioning and Covering Problems

Set Covering Problem and Set Partitioning Problem are models for many important industrial applications. In this paper, we solve some Operational Research benchmarks with Ant Colony Optimization using a new transition rule. A Lookahead mechanism was incorporated to check constraint consistency in each iteration. Computational results are presented showing the advantages to use this additional mechanism to Ant Colony Optimization

Using local search for guiding enumeration in constraint solving

In Constraint Programming, enumeration strategies (selection of a variable and a value of its domain) are crucial for resolution performances. We propose to use Local Search for guiding enumeration: we extend the common variable selection strategies of constraint programming and we achieve the value selection based on a Local Search. The experimental results are rather promising.

Using a Choice Function for Guiding Enumeration in Constraint Solving

In Constraint Programming, selection of a variable and a value of its domain enumeration strategies are crucial for resolution performances. We propose to use a Choice Function for guiding enumeration: we exploit search process features to dynamically adapt a Constraint Programming solver in order to more efficiently solve Constraint Satisfaction Problems. The Choice Function provides guidance to the solver by indicating which enumeration strategy should be applied next based upon the information of the search process, it should be captured through some indicators. The Choice Function is defined as a weighted sum of indicators expressing the recent improvement produced by the enumeration strategy had been called. The weights are determined by a Genetic Algorithm in a multilevel approach. We report results where our combination of strategies outperforms the use of individual strategies

A Hyperheuristic Approach for Constraint Solving

In this work we propose a Choice Function for guiding Constraint Programming in the resolution of Constraint Satisfaction Problems. We exploit some search process features to select on the fly the Enumeration Strategy (Variable + Value Selection Heuristics) in order to more efficiently solve the problem at hand. The main novelty of our approach is that we reconfigure the search based solely on performance data gathered while solving the current problem. We report encouraging results where our combination of strategies outperforms the use of individual strategies.

A Hyperheuristic Approach to Select Enumeration Strategies in Constraint Programming

This work is focused on the Enumeration phase of Constraint Programming to solve Constraint Satisfaction Problems, an enumeration strategy is constituted by a variable selection heuristic and a value selection heuristic. A suitable definition and use of the enumeration strategy can strongly improve the resolution process. In order to select the enumeration stretegies dynamically here we present a hyperheuristic approach using information about the search process. We report encouraging results where our combination of strategies outperforms the use of individual strategies

A Cultural Algorithm for Solving the Set Covering Problem

In this paper we apply a new evolutive approach for solving the Set Covering Problem. This problem is a reasonably well known NP-complete optimization problem with many real world applications. We use a Cultural Evolutionary Architecture to maintain knowledge of Diversity and Fitness learned over each generation during the search process. Our results indicate that the approach is able to produce competitive results in compare with other approximation algorithms solving a portfolio of test problems taken from the ORLIB.

A Two-Phase Set Partitioning Model for Passenger Transportation

A two-phase set partitioning model is proposed to solve passenger transportation, this problem is a real life application for the pickup and delivery problem where vehicles have to transport passengers from their locations to different destinations with minimal routing cost. Following the classical cluster-first route-second approach the problem is modelled in two phases and solved like a set partitioning problem using ant computing.