Papers:
[1]
Cover chapter 11 on tractable language in
Constraint Processing.
[2]
Cover chapter 13 in Constraint Processing.
[3] PDF
E. Gelle and B. Faltings. "Solving Mixed and Conditional Constraint
Satisfaction Problems". Constraints, Vol 8 Num 2, April 2003.
[4] PS
Bacchus et. al. "DPLL with caching: a new algorithm for #SAT and Bayesian
inference".
[5] PS
Larrosa et. al. "Solving still-life problems".
[6] PS
van Beek et. al. "A fast and simple algorithm for bounds consistency
of the alldifferent constraint". IJCAI2003, pp 245-250 .
[7]
Chmeiss and Jegou and Keddar. "On a generalization of triangulated graphs
for domain decompositions of CSPs". IJCAI 2003, pp. 203-208.
[8] PDF
Jegou and Terrioux. "Hybrid backtracking bounded by tree-decomposition
of constraint networks". AI Journal vol 146, no 1, 2003.
[9]
Terrioux and Jegou. "Bounded backtracking for the valued constraint
satisfaction problem". CP2003.
[10] PS |
PDF
Rish, I., and
Dechter, R., "Resolution versus Search: Two Strategies for SAT" In
"Journal of Automated Reasoning", special issue on SAT, Volume 24, Issue 1/2,
pp. 225-275, Januray, 2000.
[11] PS
Y. Zhang and R. Ypa, "Consistency and set intersection"
In "IJCAI-2003", pp. 263-267.
Programming Project:
Experiment with REES:
1. Implement your own solver, plug it into REES and compare.
2. Implement a new structure-based random network and compare existing
algorithms in REES.
3. For the task of counting solutions, perform an experimental analysis
using REES.
4. Develop a benchmark in a dmain of your choice using REES and run your
algorithms.
Specific ideas:
1. Implement backtracking with forward checking using generalized arc-consistency.
2. Implement variants of SLS + cycle-cutset.
3. Compare PC look-ahead with recent heuristics developed by Kalev.
4. Implement path-consisteny for look-ahead but record results on domains
only.
