Table 5 Descriptive statistics of the input and target variables used in the ML models before outlier removal consideration.

From: An efficient machine learning approach for predicting concrete chloride resistance using a comprehensive dataset

 

Variables

Count

Mean

Min

Max

Std. Dev

25%

50%

75%

Input Parameters

Water (kg/m3)

1073

165.15

8.46

1049.39

96.00

122.50

168.00

189.00

Cement (kg/m3)

1073

340.23

13.02

2384.97

189.34

250.96

344.10

414.00

Slag (kg/m3)

1073

42.40

0.00

1284.44

113.99

0.00

0.00

83.65

Fly ash (kg/m3)

1073

34.79

0.00

735.00

116.06

0.00

0.00

0.00

Silica fume (kg/m3)

1073

7.68

0.00

468.50

31.19

0.00

0.00

0.00

Fine aggregate (kg/m3)

1073

755.99

0.00

1574. 10

236.82

688.79

769.48

835.00

Coarse aggregate (kg/m3)

1073

846.61

0.00

1704.00

325.52

607.00

957.00

1064.93

Superplasticizer (kg/m3)

1073

1.53

0.00

10 20

1.94

0.00

0.00

3.00

Fresh density (kg/m3)

1073

2211.57

1364.00

2467.00

234.15

2162.00

2297.40

2379.90

Comp. str. test age (days)

1073

28.54

7.00

180.00

16.16

28.00

28.00

28.00

Compressive strength (MPa)

1073

46.62

18.90

80 00

11.63

38.99

46.37

54.60

Migration test age (days)

1073

120.07

3.00

2880.00

264.77

28 00

28.00

91.00

Output

Dnssm (× 10−12 m2/s)

1073

8.010

0.220

133.600

8.496

3.466

6.600

9.679