### Table 1 specifies the set of possible transitions that can take place in the

### TABLE IV CLASSIFICATION ACCURACIES WITH STATE, OBSERVATION OR ALL COVARIANCES CONSTRAINED TO BE DIAGONAL, WITH H SPECIFIED OR SET TO THE IDENTITY MATRIX.

### Table 16: DoE set and expert specified target output

2001

"... In PAGE 11: ... Even for this very constrained problem, 729 was a number too large for evaluation and hence had to be reduced. This was accomplished by performing a half factorial DoE (Fowlkes and Creveling, 1995) which cuts that number to 45 experiments as can be seen in Table16 . The result of the expert evaluation is seen in the columns labeled Target 1 and Target 2 which represent the diagnosis for Fault 1 and Fault 2, respectively.... ..."

Cited by 6

### Table 1: The standard set of visemes specified in MPEG-4 and related phonemes

2006

### Table 3. COPS model statistics.

"... In PAGE 9: ... This resulted in a total of 64 models in the benchmark test set. Table3 gives an overview of the model sizes. For full model statistics information, see ref.... ..."

### Table 4 can also be specified as permutations on sets of 4 elements, as shown in Table 5. The

2000

"... In PAGE 9: ... The mux instruction in IA-64 [15] and the vperm instruction in Altivec [16] are similar. Table4 gives examples of this permute instruction on 8-bit and 16-bit subwords. Table 4: Examples of Permute Instruction on 8-bit and 16-bit Subwords Operand register: R1 = a b c d e f g h permute Instruction Result register contents Type of Permutation permute,8,01234567 R1, Rt Rt = a b c d e f g h identity permutation permute,8,10325476 R1, Rt Rt = b a d c f e h g exchange permute,8,66666666 R1, Rt Rt = g g g g g g g g broadcast permute,8,76543210 R1, Rt Rt = h g f e d c b a reverse permute,8,05276341 R1, Rt Rt = a f c h g d e b arbitrary permutation permute,8,55000366 R1, Rt Rt = f f a a a d g g permutation with repetitions permute,16,0213 R1, Rt Rt = a b e f c d g h permuting four 16-bit subwords There is a limit to the efficiency of the permute instruction for permuting many subwords, since the control bits quickly exceed the number of bits permuted.... In PAGE 9: ... Table 4 gives examples of this permute instruction on 8-bit and 16-bit subwords. Table4 : Examples of Permute Instruction on 8-bit and 16-bit Subwords Operand register: R1 = a b c d e f g h permute Instruction Result register contents Type of Permutation permute,8,01234567 R1, Rt Rt = a b c d e f g h identity permutation permute,8,10325476 R1, Rt Rt = b a d c f e h g exchange permute,8,66666666 R1, Rt Rt = g g g g g g g g broadcast permute,8,76543210 R1, Rt Rt = h g f e d c b a reverse permute,8,05276341 R1, Rt Rt = a f c h g d e b arbitrary permutation permute,8,55000366 R1, Rt Rt = f f a a a d g g permutation with repetitions permute,16,0213 R1, Rt Rt = a b e f c d g h permuting four 16-bit subwords There is a limit to the efficiency of the permute instruction for permuting many subwords, since the control bits quickly exceed the number of bits permuted. Permuting four subwords requires only 8 control bits, which can be encoded in the permute instruction itself [4, 15].... In PAGE 10: ...Table4... ..."

Cited by 14

### Table 2: Results from combining skills and help, where the agent has the minimal skill set plus the specified skill.

"... In PAGE 3: ... However, we predicted that the pro- cess to reach solutions would improve and that this would be reflected in gains of energy remaining. Results Consistent with our prediction, Table2 shows very little change in success when we add a single skill to an agent and provide help all the while. When averaged altogether, the success rate over all the worlds and all the skills went up by a single percentage point from 48 to 49 percent.... ..."

### Table 1. The number of successful trials where the optimal rule set was found during the specified number of generations.

"... In PAGE 31: ... For comparing the search ability of our GBML algorithm with/without the four heuristics, we examined when the optimal rule set of the three fuzzy if-then rules in (2)-(4) was found during the execution of our GBML algorithm in each trial. Table1 shows the relation between the number of generations and the number of successful trials where the optimal rule set was found. For example, the optimal rule set was found in 12 out of 20 trials during the first 2000 generations when we used the four heuristics.... In PAGE 31: ... For example, the optimal rule set was found in 12 out of 20 trials during the first 2000 generations when we used the four heuristics. From Table1 , we can see that the use of the four heuristics significantly improved the search ability of our GBML algorithm. Table 1.... ..."

### Table 1 shows three example constraints. Place-in-region con- straints specify in which region of the disk a set of blocks should be placed.1 Place-seq constraints specify a set of blocks that should be placed sequentially. Place-as-set constraints specify a set of blocks that should be placed adjacent and fetched as a single unit. These three constraint types have been sufficient for many heuristics, but additions are expected for some future heuristics. For example, heuristics that exploit replication and low-level device characteris- tics (e.g. track-aligned extents [22]) may require additions.

2003

Cited by 9