Tools
Sorted by:
Try your query at:
 |
 |
 |
 |
 |
 |
Results 1 - 10
of
22,137
Table VI. comparison of the estimation without median constraints with previous results.
1999
Cited by 9
Table 2: Average, minimum, maximum sensitivity and specificity over tRNA and 5S RNA alignments in RFAM database for the proposed probabilistic alignment constraint (New constraint) and the previously employed M constraint.
2006
"... In PAGE 8: ...imilarity, e.g. by structural alignments. This, in part, lim- its the sensitivity that may be obtained using most meth- ods based on sequence nucleotide similarity alone (including ours). Table2 summarizes the average values of sensitivity and specificity over tRNA and 5S RNA alignments in RFAM database for the proposed probabilistic alignment con- straint, the previously employed M constraint, and for sin- gle sequence prediction [19]. Once again, the average values indicate the superior performance of the proposed method over the previous M constraint.... ..."
Table 5: Average, minimum, maximum run times (in seconds) and memory (in megabytes) requirement results of proposed probabilistic alignment constraint (New constraint) and the previously employed M constraint.
2006
"... In PAGE 10: ... Since these require significantly more time than the shorter tRNA sequences the overall impact of the speed-up is very significant and in fact increases the length of sequences on which Dynalign can be deployed. The memory requirements for the two methods are com- pared in Table5 , where the minimum, maximum, and average memory (in megabytes) required for the 100 sequence pairs each of tRNAs and 5S RNAs are indicated. Memory requirements are as reported in the size entry of Linux ps command after all requisite dynamic allocations are done.... ..."
Table 2. Parameters and models used in some previous studies of random constraint satisfaction problems.
Tableau 2: Optimization in fragment (5); * = constraint violation; ok=constraint satisfaction; =currently preferred interpretation; =previously preferred interpretation; S=Stay; UT=Unique Topic; FD=Forward Discourse.
Tableau 3: Optimization in fragment (6); * = constraint violation; ok=constraint satisfaction; =currently preferred interpretation; =previously preferred interpretation; S=Stay; UT=Unique Topic; FD=Forward Discourse.
Table 1. Parameters and models used in some previous studies of random constraint satisfaction problems. The nal column details studies in which model B or C was used and p2 lt; 1=m. In the limit, such problem classes are not trivially insoluble.
Table 2 presents the results of iteratively building up constraints to remove symmetric paths. The experiment was performed with and without pairwise ordering constraints to gauge their utility in removing symmetric paths. At each depth constraints generated at all previous depths are used to minimise the number of extra constraints needed. Even with this incremental approach and the ordering constraints, a substantial set of new constraints is necessary to break the path symmetry by depth 7. Discovering and adding constraints deeper in the search tree is unfortunately not currently feasible. Preliminary experiments also suggest that such a shallow part of the tree is not visited often enough for the overhead of these constraints to be worthwhile.
"... In PAGE 8: ... Table2... ..."
TABLE I PREVIOUS WORK SEEMS TO IMPLY THAT THE LOGICAL IMPLICATION OF CI AND EMVD COINCIDES AS LONG AS A RESTRICTION IS IMPOSED ON THE INPUT SET OF CONSTRAINTS.
TABLE I PREVIOUS WORK SEEMS TO IMPLY THAT THE LOGICAL IMPLICATION OF CI AND EMVD COINCIDES AS LONG AS A RESTRICTION IS IMPOSED ON THE INPUT SET OF CONSTRAINTS.
Results 1 - 10
of
22,137
