Table 2 Evaluation metrices of machine learning models employed for risk prediction of HPAI.
Sl. No | Models | Model Specification | KAPPA | ROC | TSS | AUC | Accuracy | ERROR RATE | F1 SCORE | LOG LOSS |
|---|---|---|---|---|---|---|---|---|---|---|
1. | GLM | \(\:E\left(Y\right|X)=\mu\:={g}^{-1}(X\beta\:)\) Y- Expected Value, X-Conditional, \(\:\:X\beta\:\)- Linear Predicator, g-Link Function | 0.34 | 0.80 | 0.49 | 0.80 | 0.75 | 0.25 | 0.89 | 0.43 |
2. | GAM | \(\:g\left(E\left(Y\right)\right)={\beta\:}_{0}+{f}_{1}\left({x}_{1}\right)+{f}_{2}\left({x}_{2}\right)+\cdots+{f}_{i}\left({x}_{i}\right)\) Y-Response Variable, g-Link Function, fi-Specified Parametric Form, xi-Predicator Variable | 0.34 | 0.80 | 0.49 | 0.80 | 0.75 | 0.25 | 0.89 | 0.43 |
3. | RF | Y = \(\:\sum\:_{i=1}^{n}\text{f}\left(\text{t}\text{n}\right)\) Y- Average of aggregated predictions of the multiple decision trees, tn – multiple decision trees trained on different subset of the same training data | 0.68 | 1.00 | 0.96 | 1.00 | 0.98 | 0.02 | 0.99 | 0.12 |
4. | GBM | \(\:f\left(x\right)=arg\:mi{n}_{\theta\:}\sum\:_{i=1}^{n}L({y}_{i}\:,\theta\:)+\sum\:_{m=1}^{M}\eta\:{\rho\:}_{m}{\varphi\:}_{m}\left(x\right)\) m- Iteration, \(\:\:\eta\:\)-Learning Rate, \(\:{\rho\:}_{m}\)- Step length | 0.50 | 0.93 | 0.72 | 0.93 | 0.84 | 0.16 | 0.92 | 0.31 |
5. | NNET | \(\:Y=f\left(\sum\:_{i=1}^{n}{x}_{i}{w}_{i}\right)+b\) Y-Output, \(\:{x}_{i}\)-Inputs, \(\:{w}_{i}\)- Weights, \(\:\:b\)- Bias | 0.00 | 0.50 | 0.00 | 0.50 | 0.25 | 0.75 | 0.00 | 25.90 |
6. | MARS | \(\:\widehat{f}\left(x\right)=\sum\:_{i=1}^{k}{c}_{i}{B}_{i}\left(x\right)\) \(\:{c}_{i}\)- Constant Coefficient, \(\:{B}_{i}\left(x\right)\)- Basis Function | 0.41 | 0.87 | 0.60 | 0.87 | 0.80 | 0.21 | 0.91 | 0.34 |
7. | FDA | \(\:{\eta\:}_{l}\left(x\right)={X}^{T}{\beta\:}_{l}\) | -0.01 | 0.50 | -0.01 | 0.50 | 0.79 | 0.21 | 0.88 | 7.28 |
8. | CT | \(\:f\left(x\right)=\sum\:_{j=1}^{T}{w}_{j}I(x\in\:{R}_{j})\) | 0.66 | 0.93 | 0.74 | 0.93 | 0.85 | 0.15 | 0.93 | 0.34 |
9. | SVM | \(\:\{x:f(x)={x}^{T}\beta\:+{\beta\:}_{0}=0\}\) | 0.51 | 0.87 | 0.68 | 0.87 | 0.85 | 0.15 | 0.91 | 0.53 |
10. | NB | \(\:P\left(c\right|x)=\frac{P\left(x\right|c\left)P\right(c)}{P\left(x\right)}\) \(\:P\left(c\right|x)\)-Posterior Probability \(\:\:P(x\left|c\right)\)-Likelihood, \(\:P\left(c\right)\)-Class Prior Probability, \(\:\:P\left(c\right)\)-Predictor Prior Probability | -0.20 | 0.72 | -0.18 | 0.72 | 0.30 | 0.70 | 0.47 | 3.55 |
11. | ADA | \(\:{F}_{T}\left(x\right)=\sum\:_{t=1}^{T}{f}_{t}\left(x\right)\) \(\:{f}_{t}\)- Weak Learner, \(\:\:x\)- Input,\(\:\:\:T\)- \(\:\:T\)th Positive or Negative Classifier | 0.73 | 0.83 | 0.66 | 0.83 | 0.92 | 0.08 | 0.95 | 2.71 |
