Table 3 5-class classification: wastewater without CRP, \(10^{-4} \, \upmu\)g/ml, \(10^{-3} \, \upmu\)g/ml, \(10^{-2} \, \upmu\)g/ml, and \(10^{-1} \, \upmu\)g/ml CRP (+400 nm spectrum). The reported values represent the average performance across 30 repetitions. The best value for each metric, highlighted in bold, corresponds to the single best-performing repetition among all trials.

Cubic SVM

63.3532

63.3168

63.2059

63.2612

90.8444

Fine Gaussian SVM

59.3492

59.5536

59.2146

59.3835

89.8426

Wide NN

57.4206

57.4147

57.2792

57.3468

89.3640

Mediumk NN

55.7738

55.7569

55.6309

55.6937

88.9518

Fine KNN

55.9127

56.1932

55.8686

56.0303

88.9809

Weighted KNN

54.3294

54.5295

54.2231

54.3757

88.5824

Cosine KNN

53.7897

53.5448

53.6330

53.5888

88.4496

Medium Gaussian SVM

53.5556

56.2935

53.4102

54.8125

88.3999

NArrow NN

51.8254

51.8435

51.6785

51.7607

87.9654

Bilayered NN

51.4127

51.3504

51.2731

51.3112

87.8609

Linear Discriminant

51.3968

50.8151

51.1057

50.9599

87.8571

Cubic KNN

51.1230

51.4212

50.9866

51.2027

87.7823

Bagged Trees (Ensemble)

50.7619

50.6356

50.6067

50.6211

87.6863

Trilayered NN

50.3611

50.2789

50.2067

50.2425

87.5984

Medium KNN

50.1825

50.3821

50.0457

50.2132

87.5479

Coarse Gaussian SVM

49.0159

51.6679

48.7692

50.1755

87.2443

Linear SVM

48.0595

48.5859

47.8147

48.1967

87.0181

Fine Tree

38.0357

38.1712

37.9857

38.0779

84.5114

RUSBoosted Trees (Ensemble)

36.9524

38.8702

36.8793

37.8441

84.2478

Medium Tree

34.9762

37.4039

34.8757

36.0884

83.7449

Kernel Naive Bayes

33.2143

35.6113

33.1137

34.3160

83.2970

Subspace KNN (Ensemble)

28.4405

32.4879

27.8565

29.9719

81.9716

Subspace Discriminant (Ensemble)

27.7659

31.7558

27.1160

29.1964

81.8153

Gaussian Naive Bayes

24.0635

32.1147

24.3561

27.6863

81.0744

Best result (CSVM)

64.8810

64.8072

64.7473

64.7773

91.2273