Table 3 Different ML classifiers and their prediction parameters for the raw impedance data for six bacteria culture.
Classifier | Bacteria type | Precision | Recall | F1-score | Classification Accuracy |
|---|---|---|---|---|---|
Random Forest | B.C. | 0.44 | 0.47 | 0.45 | 71% |
S.A. | 0.92 | 0.91 | 0.91 | ||
P.F. | 0.59 | 0.54 | 0.57 | ||
E. coli | 1 | 1 | 1 | ||
L.P. | 0.84 | 0.78 | 0.81 | ||
B.S. | 0.51 | 0.57 | 0.53 | ||
Support Vector Machine | B.C. | 0.06 | 0.01 | 0.01 | 20% |
S.A. | 0 | 0 | 0 | ||
P.F. | 0 | 0 | 0 | ||
E. coli | 0.86 | 0.29 | 0.43 | ||
L.P. | 0.15 | 0.08 | 0.1 | ||
B.S. | 0.17 | 0.9 | 0.28 | ||
K Nearest Neighbours | B.C. | 0.19 | 0.37 | 0.25 | 35% |
S.A. | 0.43 | 0.42 | 0.42 | ||
P.F. | 0.24 | 0.23 | 0.24 | ||
E. coli | 0.95 | 0.62 | 0.75 | ||
L.P. | 0.41 | 0.21 | 0.27 | ||
B.S. | 0.29 | 0.27 | 0.28 | ||
Decision Tree | B.C. | 0.55 | 0.56 | 0.55 | 77% |
S.A. | 0.92 | 0.91 | 0.91 | ||
P.F. | 0.68 | 0.65 | 0.66 | ||
E. coli | 1 | 1 | 1 | ||
L.P. | 0.85 | 0.81 | 0.83 | ||
B.S. | 0.64 | 0.69 | 0.66 | ||
Gradient Boost Accuracy | B.C. | 0.28 | 0.26 | 0.27 | 67% |
S.A. | 0.43 | 0.13 | 0.2 | ||
P.F. | 0.31 | 0.28 | 0.29 | ||
E. coli | 1 | 0.49 | 0.66 | ||
L.P. | 0.23 | 0.42 | 0.29 | ||
B.S. | 0.3 | 0.46 | 0.36 | ||
Bagging Classifier | B.C. | 0.69 | 0.7 | 0.69 | 83% |
S.A. | 0.94 | 0.95 | 0.94 | ||
P.F. | 0.73 | 0.73 | 0.73 | ||
E. coli | 1 | 1 | 1 | ||
L.P. | 0.93 | 0.94 | 0.93 | ||
B.S. | 0.71 | 0.72 | 0.72 | ||
Gaussian Naive Bayes | B.C. | 0.52 | 0.18 | 0.27 | 26% |
S.A. | 0.19 | 0.68 | 0.3 | ||
P.F. | 0.08 | 0.01 | 0.02 | ||
E. coli | 0.93 | 0.25 | 0.4 | ||
L.P. | 0.27 | 0.34 | 0.3 | ||
B.S. | 0.21 | 0.06 | 0.1 | ||
Multilayer Perceptron Neural Network | B.C. | 0.42 | 0.26 | 0.32 | 34% |
S.A. | 0.27 | 0.58 | 0.37 | ||
P.F. | 0.5 | 0.09 | 0.15 | ||
E. coli | 0.91 | 0.37 | 0.53 | ||
L.P. | 0.36 | 0.32 | 0.34 | ||
B.S. | 0.25 | 0.46 | 0.32 |
