### Table 1. Problem instances

"... In PAGE 8: ... Due to the limited space here, we only show the results for the throughput maximization case. There are in total six problem instances, A1, A2, A3, B1, B2 and B3 as shown in Table1 . Problems A and problems B are different in the system configurations: the number... ..."

### Table 1 Problem Instances

"... In PAGE 6: ... The algorithms presented in this paper are tested on ten instances which vary in the number of nurses, cover requirements, shift types, constraint types and priorities, personal requests and planning horizon. Table1 provides some more information on the instances. Instance Nurses Shift types Skill levels Planning horizon BCV-1.... ..."

Cited by 1

### Table 1 Problem Instances

"... In PAGE 47: ... Similarily an \at end quot; { \at end quot; con ict will lead to an earliest end- ing time for Oi at e(Oj)+ , expressing that the ending times are o by . Table1 lists m l (Oj; Oi) for all pos- sible values of m; l 2 M. If there is more than one con ict for one operator pair, we have to compute the maximum value derived for the individual con icts.... In PAGE 57: ...ormance. CRIKEY does perform quicker on problem 10. This is because whereas the FF System resorts to planning from the initial state if it finds a dead end, CRIKEY will start complete search from where it entered the plateau which led to the dead end. Table1 : DriverLog SimpleTime Times pfile FFSystem CRIKEY SAPA 2 620 1150 7500 4 600 1880 8550 6 740 5990 2080 8 2580 14350 2180 10 2510 840 1560 Table 2: DriverLog SimpleTime Plan Quality pfile FFSystem CRIKEY SAPA 2 100.03 126.... In PAGE 106: ... The reasoning mechanisms underlying the pro- posed planner, that we call CPT, yield a solution to this prob- lem by pure inference and no search. This is remarkable as the inferences are not trivial and existing optimal plan- ners do not scale up well in this domain (see Table1 ). How does CPT do it? First, it infers that each subgoal on(bi; bi+1) must be achieved by the action stack(bi; bi+1) and then that these actions must be ordered sequentially, stack(bn 1; bn) rst, then stack(bn 2; bn 1), and so on.... In PAGE 106: ...70 - - - 26 tower-15 13.65 - - - 28 Table1 : Results for TOWER-n domain Table 1 shows results for CPT in relation to other three modern planners: two optimal parallel planners, Blackbox (with Chaff) (Kautz amp; Selman 1999) and IPP (Koehler et al. 1997), and an optimal temporal planner TP4 (Haslum amp; Geffner 2001).... In PAGE 106: ...70 - - - 26 tower-15 13.65 - - - 28 Table 1: Results for TOWER-n domain Table1 shows results for CPT in relation to other three modern planners: two optimal parallel planners, Blackbox (with Chaff) (Kautz amp; Selman 1999) and IPP (Koehler et al. 1997), and an optimal temporal planner TP4 (Haslum amp; Geffner 2001).... ..."

### Table 1: Problem instances

### Table 7: Notation and Problem Data for Table 6 N Number of Operations Z

### Table 10: Larger problem instances

"... In PAGE 18: ...3 Larger problems In order to study the performance of our algorithm in a more comprehensive fashion, we also carried out tests involving fifty data sets arising from large, dense networks. Table10 describes the five larger problem instances that we generated. The numerical parameters for these runs were chosen using the same recipe as for the medium-sized problems; but since these larger problem instances were likely to be far more demanding problem instances, we only used the triple A = 50000, epsilon1 = 1.... ..."

### Table I:Problem instances Problem instance Description

### Table 1. Characteristics of problem instances

"... In PAGE 9: ... 5 Computational Experiments and Discussion Experiments were done in the set of instances originated from [14], and the data referred to public Brazilian high schools, with 25 lesson periods per week for each class, in di erent shifts. In Table1 some of the characteristics of the instances can be veri ed, such as dimension and sparseness ratio (sr), which can be computed considering the total number of lessons (#lessons) and the total number of unavailable periods (#u): sr = t p (#lessons+#u) t p . Lower sparseness values indicate more restrictive problems and, likewise, problems in which it is... ..."

### Table 3 { larger problem instances

1997

"... In PAGE 4: ... For these examples the exact algorithms fail due to their exponential behavior. The results are given in Table3 . As can been seen PQ is much faster and ad- ditionally obtains smaller average test costs.... ..."

Cited by 2