Joao P. Marques-Silva. The impact of branching heuristics in propositional satisability algorithms, Proceedings of the 9th Portuguese Conference on Artificial Intelligence, 62--74, 1999.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....x with the largest combined value JW(x) JW( x) and assigns v(x) 1 if JW(x) JW( x) and v(x) 0 otherwise. Literal count heuristics: The next three branching rules are so called literal count heuristics, since only the numbers of occurences of literals in unresolved clauses are considered [17]. They make use of the following functions for a variable x: UC p (x) jfc j x 2 c and c is unresolvedgj UC n (x) jfc j :x 2 c and c is unresolvedgj DLCS: Dynamic Largest Combined Sum (DLCS) selects the variable x with largest value UC p (x) UC n (x) and assigns v(x) 1 if UC p (x) UC n ....

....are average values of 30 experiments. RDLIS: To balance the greedy behaviour of DLIS, Random DLIS selects the same variable as DLIS, but makes a random assignment v(x) RAND: RAND selects randomly an unassigned variable x and assigns randomly 0 or 1 to x. 5 Comparison of branching rules In [17], a detailed comparison between the branching rules mentioned above is presented. To validate these results we performed experiments for the same set of DIMACS benchmarks (see Table 1) on a AMD Athlon(TM) XP1700 , restricted to 512MB main memory and 180sec of CPU runtime for each instance. Each ....

J.P. Marques-Silva. The impact of branching heuristics in propositional satisfiability algorithms. In 9th Portuguese Conference on Artificial Intelligence (EPIA), 1999.

....is unsatisfiable. Readers are referred to [6] for a more detailed discussion of the DPLL algorithm. There are a large number of different SAT solvers that differ mainly in how each of these functions is implemented using different heuristics. A lot of effort has been spent on decisionmaking (e.g. [13][14] 9] and significant progress has been made on how efficient deduction (e.g. 9] 15] 8] However, to our knowledge, only [7] and [6] and its variations) have discussed implementation of conflict driven learning. The authors are not aware of any evaluation of different conflict driven ....

J. P. Marques-Silva, "The Impact of Branching Heuristics in Propositional Satisfiability Algorithms," Proceedings of the 9th Portuguese Conference on Artificial Intelligence (EPIA), September 1999.

.... basic DP procedure: 2 A lot of branching rules were proposed [5, 7, 4] each aiming at accelerating the search under different preferences (e.g. eliminating small clauses, weighting variables that occur often as positive as well as negative literal, A good overview and a comparison gives [6]. Most branching rules are variable state dependent, i.e. statistics on the occurence of unnassigned variables are used whereby the underlying clause database is assumed to be simplified using the already assigned variables. Davis Putnam( CNF f ) f if f is empty return SATISFIABLE if 2 2 f ....

J.P. Marques-Silva. The impact of branching heuristics in propositional satisfiability algorithms. In 9th Portuguese Conference on Artificial Intelligence (EPIA), 1999.

....branching rule selects variable x with largest combined value JW(x) JW( x) and assigns v(x) 1 if JW(x) JW( x) and v(x) 0 otherwise. The next three branching rules are so called literal count heuristics, since only the numbers of occurences of literals in unresolved clauses are considered [13]. They make use of the following functions for a variable x: UC p (x) jfc j x 2 c and c is unresolvedgj UC n (x) jfc j :x 2 c and c is unresolvedgj DLCS Dynamic Largest Combined Sum (DLCS) selects the variable x with largest value UC p (x) UC n (x) and assigns v(x) 1 if UC p (x) UC ....

....default branching rule used in GRASP. RDLIS To balance the greedy behaviour of DLIS, Random DLIS selects the same variable as DLIS, but makes a random assignment v(x) RAND RAND selects randomly an unassigned variable x and assigns randomly 0 or 1 to x. 5. COMPARISON OF BRANCHING RULES In [13], a detailed comparison between the braching rules mentioned above is presented. To validate this results we perfomed experiments for the same set of DIMACS benchmarks (see Table 1) on a SUN Sparc Ultra4 with 248Mhz, restricted to 512MB main memory and 2h of CPU runtime for each instance. For each ....

J.P. Marques-Silva. The impact of branching heuristics in propositional satisfiability algorithms. In 9th Portuguese Conference on Artificial Intelligence (EPIA), 1999.

No context found.

Joao P. Marques-Silva. The impact of branching heuristics in propositional satisability algorithms, Proceedings of the 9th Portuguese Conference on Artificial Intelligence, 62--74, 1999.

No context found.

J. P. Marques-Silva. The impact of branching heuristics in propositional satis ability algorithms. In P. Barahona and J. Alferes, editors, Proceedings of the 9th Portuguese Conference on Arti cial Intelligence, LNAI 1695, pages 62-74. SpringerVerlag, September 1999.

No context found.

J. P. Marques-Silva, "The impact of branching heuristics in propositional satisfiability algorithms," 9th Portugese Conference on Artificial Intelligence, September, 1999.

No context found.

J. P. Marques-Silva, "The impact of branching heuristics in propositional satisfiability algorithms," 9th Portugese Conference on Artificial Intelligence, September, 1999.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC