### Table 2 Algorithms for graph matching

1999

"... In PAGE 10: ... The experimental conditions are summarised in Table 2. Each of the algorithms listed in Table2 , except HC, was run 100 times. Since HC is deterministic, it was only run once per graph.... ..."

### Table 4: Graph theoretic classes for program visualizations

"... In PAGE 55: ... The review criteria were applied to each diagram appropriately, in terms of its graph-theoretic class, conformance, presentation and interaction. A summary of the results of applying graph theoretic class criteria to modelling diagrams is in Table 3 and program visualizations is in Table4 . Each diagram in the review is described by its graph type(s) and the four boolean attributes: directed, cyclic, connected and planar.... ..."

Cited by 1

### Table 1: Experimental parameter ranges for the genetic algorithm.

1996

"... In PAGE 9: ... per treatment. The parameter ranges used for each circuit are shown in Table1 for the genetic algorithm, and in Table 2 for the simulated annealing algorithm. For each graph, the mean cutsizes of the genetic algorithm and simulated annealing are compared.... ..."

Cited by 1

### Table 3. Results for 144X144 and 64X64 Window Size: 2D search and Genetic Algorithm

2000

"... In PAGE 10: ... For ease of description, the results of first 10 pairs are included in this result comparison section. Table3 lists the results for 2D search and genetic algorithm, when window size of 144X144 (larger window) and 64X64 (smaller window) are used. Table 4 lists the same result when window size of 196X196 (larger window) and 64X64 (smaller window) are used.... ..."

Cited by 2

### Table 2. Parameters of the genetic algorithm for load balancing

2005

"... In PAGE 7: ... GALB: Using the developed genetic algorithm which performs load balancing with the described initialization GALBR: Same as GALB but using random initialization of the genetic algorithm RD: Random allocation of lots to appropriate machines The algorithm was developed using Visual C++ and the testing was performed in a Windows XP environment on a PC Pentium 2 GHz with 512 MB of RAM. Parame- ters of the GA are given in Table2 . The algorithm was run 5 times and the results ob- tained using real-world data for one month planning horizon discretised homogene- ously into one minute unit time periods, are given in Table 3.... ..."

Cited by 1

### Table 1 Genetic-algorithm encoding of a multilayer neural network spatial interaction model Bits Meaning

1998

"... In PAGE 16: ... On the other hand, using a larger quantity of data to evaluate the CNN may imply that bigger networks could be trained more precisely than smaller ones and, thus, the implicit pruning process would be reluctant to remove links. Encoding Scheme: Table1 illustrates how a string is built. The string representation has several desirable properties.... ..."

### Table 1: Sample frame rates for the visualization algorithm.

2003

"... In PAGE 7: ...2 with a 1 GHz Pentium III dual processor and 1 GB of RAM. The performance times reported in Table1 support interactive exploration of un- steady flow on surfaces. The first time reported in the FPS column is for the static cases of steady-state visualization and the absence of changes to the view point.... In PAGE 7: ... More specifically, the dynamic cases require the con- struction of a velocity image, image overlay, as well as geometric edge detection. We include geometric edge detection in the frame rates reported in Table1 . It does not introduce significant overhead since it is easily built into the advection process itself.... ..."

Cited by 34

### Table 1. Upper Bounds for 2-D Orthogonal Graph Drawing

1999

"... In PAGE 3: ... Using a diagonal layout our algorithm produces 2-degree-restricted square-drawings. Table1 summarizes bounds for 2-D orthogonal graph drawing. Table 1.... ..."

Cited by 19

### Table 2(d): Average NSLs for irregular task graphs of various sizes.

1997

"... In PAGE 7: ... The normalized schedule length (NSL), defined as the ratio of schedule length to the critical-path length, is computed for each solution generated by the algorithms. In Table2 (a), the average NSLs produced by each algorithm for each graph type (averaged over 7 values of CCR and 10 different graph sizes) are shown. These numbers clearly indicate that the CPFD algorithm produced the shortest average schedule length not only across all graphs types but also for Algorithm LWB LCTD DSH BTDH PY CPFD LU MVA InTree OutTree ForkJoin Random 1.... In PAGE 7: ...54 1.50 Table2 (a): Average NSLs across all graph types. Algorithm LWB LCTD DSH BTDH PY CPFD 0.... In PAGE 7: ...75 2.21 Table2 (b): Average NSLs across all CCRs. Algorithm LWB LCTD DSH BTDH PY CPFD 15 16 17 18 19 20 21 22 23 24 1.... In PAGE 7: ...35 1.36 Table2 (c):Average NSLs for regular task graphs of various matrix dimensions.... In PAGE 8: ... Based on this comparison, the ranking of the six algorithms is as follows: CPFD, BTDH, DSH, LCTD, PY, and LWB. Table2 (b) shows the NSLs (averaged over graph size and graph type) of each algorithm against various values of CCR. We can observe that all algorithms were very sensitive to the value of CCR.... In PAGE 8: ... The ranking of the algorithms based on performance, however, is consistent with our earlier conclusion. Table2 (c) and Table 2(d) show the NSLs yielded by each algorithm against various graph types (averaged across graph types and CCRs), for regular and irregular graphs, respectively. We can notice that in this context the CPFD was also consistently better than all other algorithms.... ..."

Cited by 1