Skip to main content

Table 2 Experiment data with results using all three different approaches

From: SURF: identifying and allocating resources during Out-of-Hospital Cardiac Arrest

Test case

Users

AEDs

Emergencies

Processing time (s)

Unique emergencies covered

    

Hungarian

ILP

Preprocessed ILP

Hungarian

ILP

Preprocessed ILP

1

50

2

1

0.01

0.06

0.03

1

1

1

2

100

3

1

0.05

0.08

0.01

1

1

1

3

150

4

1

0.10

0.09

0.02

1

1

1

4

200

5

1

0.24

0.07

0.02

1

1

1

5

250

6

1

0.42

0.09

0.02

1

1

1

6

300

7

1

0.74

0.16

0.03

1

1

1

7

350

8

1

1.06

0.19

0.05

1

1

1

8

400

9

1

1.53

0.24

0.05

1

1

1

9

450

10

1

2.20

0.21

0.04

1

1

1

10

500

10

1

2.55

0.23

0.05

1

1

1

11

550

10

1

2.97

0.24

0.04

1

1

1

12

600

20

1

6.95

0.48

0.06

1

1

1

13

650

30

1

13.01

0.76

0.12

1

1

1

14

700

40

1

19.66

1.21

0.16

1

1

1

15

750

50

1

34.10

1.69

0.31

1

1

1

16

800

60

2

105.18

5.76

0.35

2

2

2

17

850

70

2

146.26

7.10

0.45

2

2

2

18

900

80

2

183.78

8.84

0.54

2

2

2

19

950

90

2

222.70

9.53

0.63

2

2

2

20

1000

100

2

251.78

11.68

0.78

2

2

2

21

1050

110

3

571.15

20.10

1.08

3

3

3

22

1100

120

3

689.33

23.73

1.20

3

3

3

23

1150

130

3

843.60

26.63

1.46

3

3

3

24

1200

140

3

1010.60

32.79

1.42

3

3

3

25

1250

150

3

1112.83

41.13

1.68

3

3

3

26

1300

160

4

2178.66

61.08

2.06

4

4

4

27

1350

170

4

2359.92

67.09

2.40

4

4

4

28

1400

180

4

2969.93

75.04

3.08

4

4

4

29

1450

190

4

3329.03

80.15

2.94

4

4

4

30

1500

200

4

3671.05

83.53

3.23

4

4

4

31

1550

210

5

7682.36

120.98

4.00

5

5

5

32

1600

220

5

8442.53

128.44

4.74

5

5

5

33

1650

230

5

9353.51

130.15

4.95

5

5

5

34

1700

240

5

11,387.03

139.14

4.73

5

5

5

35

1750

250

5

11,823.76

143.14

6.04

5

5

5

\(36^{{\mathrm{a}}}\)

4

4

2

0.001

0.005

0.003

1

1

2

  1. \({}^{{\mathrm{a}}}\)A special case to show the greedy matching performed by Hungarian and ILP approaches