### Table 1: A cutting plane algorithm applied to a 120-city TSP.

1996

"... In PAGE 25: ... The optimal solution was found after 13 calls to the LP-solver. The value of the LP relaxation, z LP , and the number of added cuts at each iteration, are given in Table1 . As can be concluded from the table,... ..."

Cited by 4

### Table 1: A cutting plane algorithm applied to a 120-city TSP.

1996

"... In PAGE 23: ... The optimal solution was found after 13 calls to the LP-solver. The value of the LP relaxation, zLP , and the number of added cuts at each iteration, are given in Table1 . In total 36 subtour elimination constraints, 25 2-matching constraints, and 35 comb constraints were added.... ..."

Cited by 4

### Table 1: Summary of First Iteration Solutions for Three Heuristics

2007

"... In PAGE 12: ... The implementation was done in Matlab with the COPL LP Solver [20] and the Concorde TSP solver [6]. In Table1 and 2, we compare the initial solutions provided by three heuristics: the LP- based heuristic, the region partition heuristic and a traditional nearest-neighbor approach which can be stated as follows:... In PAGE 13: ... We have made the following observations based on the simulation results presented in the tables. First, according to Table1 and 2, both LP-based and region partition heuristics pro- duce notably better initial routes than the traditional nearest neighbor approach. They are efficient and effective candidates for the initial assignment.... ..."

Cited by 2

### Table 4: TSP

1999

"... In PAGE 10: ... Processes alternately retrieve tours from the queue, split them into sub-tours, and put them back into the queue. As shown in Table4 , TSP is almost exclusively lock- based. Locks are used to guard access to the central queue and to current minimum tour values.... ..."

Cited by 14

### Table 1. The TSP Instances

1999

"... In PAGE 7: ... 5 Experimental results The techniques described so far were applied to three TSP instances taken from the TSPLIB95 [Rei95]. The problems are listed in Table1 , as well as the values of some of the relevant parameters for the GA. The instances referred can be classi ed as Euclidean TSPs, i.... ..."

Cited by 1

### Table 2: TSP results.

2004

"... In PAGE 9: ... By combining the integer array with the job object, the number of network latencies is cut in half. Table2 shows that the number of message sent drops significantly (up to 31%) when object combining is enabled. The slight performance gain when enabling object combining on a single machine is due to caching effects as the job object and the containing array are now (always) physically closer together in memory.... ..."

Cited by 1

### Table 1: A cutting plane algorithm applied to a 120-city TSP. it is good practice to generate and add many violated inequalities at each iteration, since in 24

1996

"... In PAGE 25: ... The optimal solution was found after 13 calls to the LP-solver. The value of the LP relaxation, zLP , and the number of added cuts at each iteration, are given in Table1 . As can be concluded from the table,... ..."

Cited by 4