Table 2 Performance of clustering (DBI) and classification (accuracy, f1-score, precision, and recall) approaches and estimation performance of rate constants range using markers for each clustering scenario.

From: Clustering micropollutants and estimating rate constants of sorption and biodegradation using machine learning approaches

Process

Input feature

DBI

Accuracy

f1-score

Precision

Recall

Estimation accuracy (N = 1)

Estimation accuracy (N = 2)

Estimation accuracy (N = 3)

Aerobic

Physicochemical properties and functional groups

0.488

0.75

(0.15)

0.61

(0.24)

0.59

(0.24)

0.65

(0.23)

0.38

0.69

0.77

Biotransformation rule

0.872

0.43

(0.16)

0.32

(0.13)

0.31

(0.14)

0.33

(0.11)

0.10

0.20

0.40

Anoxic

Physicochemical properties and functional groups

0.488

0.75

(0.15)

0.61

(0.24)

0.59

(0.24)

0.65

(0.23)

0.46

0.70

0.77

Biotransformation rule

0.872

0.43

(0.16)

0.32

(0.13)

0.31

(0.14)

0.33

(0.11)

0.30

0.40

0.40

  1. The standard deviation of classification performance is mentioned in the parentheses. The estimation performance was compared with different N values, deciding the estimation range of rate constants in Eqs. (1) and (2).
Back to article page