Table 1 Model configurations with parameters.
From: Machine learning comparison for biomarker level estimation in wastewater dynamics monitoring
Model | Configuration | ||
|---|---|---|---|
FRBS system | Membership function | Number of rules | Number of inputs |
Triangular | 5 | 3 | |
SVM models | Kernel function | Function order | Classification |
Cubic SVM | Polynomial (Scale 1.7) | 3 | ECOC 1vs1 |
Quadratic SVM | Polynomial (Scale 1.7) | 2 | ECOC 1vs1 |
Medium Gaussian SVM | Gaussian (Scale 1.7) | – | ECOC 1vs1 |
Fine Gaussian SVM | Gaussian (Scale 0.43) | – | ECOC 1vs1 |
Linear SVM | Linear (Scale auto) | – | ECOC 1vs1 |
Coarse Gaussian SVM | Gaussian (Scale 6.9) | – | ECOC 1vs1 |
Ensemble models | Number of splits | Learning cycles | Classification |
Bagged Trees | 20 | 30 | Adaptive Boosting |
RUSBoosted Trees | 20 | 30 | RUSBoost |
Subspace Discriminant | – | 30 | Subspace |
Subspace KNN | – | 30 | Subspace |
Tree models | Number of splits | Split criterion | |
Fine Tree | 100 | Gini’s diversity index | |
Medium Tree | 20 | Gini’s diversity index | |
Coarse Tree | 4 | Gini’s diversity index | |
K-Nearest neighbor models | Number of neighbors | Distance | Weight |
Weighted KNN | 10 | Euclidean | Squared inverse |
Cosine KNN | 10 | Cosine | Equal |
Fine KNN | 1 | Euclidean | Equal |
Medium KNN | 10 | Euclidean | Equal |
Cubic KNN | 10 | Minkowski (exponent 3) | Equal |
Neural network models | Layer size | Activation | Iteration |
Wide Neural Network | 100 | ReLU | 1000 |
Narrow Neural Network | 10 | ReLU | 1000 |
Bilayered Neural Network | [10 10] | ReLU | 1000 |
Medium Neural Network | 25 | ReLU | 1000 |
Trilayered Neural Network | [10 10 10] | ReLU | 1000 |
Deep Neural Network | [10 10 10 10 10 10 10 10 10 10 10 10] | ReLU & Softmax | 1000 |
Naive Bayes models | Distribution | Support | |
Kernel Naive Bayes | Kernel | Unbounded | |
Gaussian Naive Bayes | Gaussian | Unbounded | |
Discriminant models | Discriminant type | ||
Quadratic Discriminant | Quadratic | ||
Linear Discriminant | Linear | ||
SVM Kernel | Learner | Iteration | Classification |
SVM | 1000 | ECOC 1vs1 | |
