Table 2. Summary of each machine learning model employed in our study, reference and reasoning for application. |
Machine
Learning
Model |
Abbreviation |
Reference |
Reasoning |
| K-Nearest-Neighbors
regressor |
KNeighbors |
(Fix and
Hodges, Jr, J. L., 1951) |
K-nearest
neighbor (KNN) models are particularly effective in regression tasks where
local data patterns are crucial, as they base predictions on the proximity
to neighboring data points. |
| Ridge |
Ridge |
(Friedman
et al., 2010) |
Linear
regression based models are beneficial in regression tasks, particularly
when regularization is required to improve model generalization and mitigate
overfitting. |
| Lasso |
Lasso |
|
| Elastic-Net |
ElasticNet |
(Friedman
et al., 2010) |
| Decision
tree |
DecisionTree |
(Breiman,
2017) |
Tree-based
models are highly effective in regression tasks, especially in scenarios
where complex, non-linear relationships or intricate interactions exist
between the input features and the target variable. |
| Extremely
randomized trees |
ExtraTree |
(Geurts
et al., 2006) |
| Random forest |
RandomForest |
(Breiman,
2001) |
| Gradient
tree boosting |
GradientBoosting |
(Friedman,
2002) |
| Adaptive
boosting |
AdaBoost |
(Freund and
Schapire, 1995) |
|
|