Research Article
Hybridization of Machine Learning Algorithms and an Empirical Regression Model for Predicting Debris-Flow-Endangered Areas
Table 4
Comparisons of performance metrics of the predictive models.
| Model | Training data | Testing data | R2 | RMSE | MAE | R2 | RMSE | MAE |
| Single predictive model | NLRM | 0.76 | 0.079 | 0.055 | 0.58 | 0.081 | 0.057 | MARS | 0.70 | 0.068 | 0.051 | 0.46 | 0.122 | 0.081 | RF | 0.91 | 0.042 | 0.031 | 0.54 | 0.085 | 0.060 | SVM | 0.63 | 0.078 | 0.052 | 0.46 | 0.099 | 0.066 |
| Hybrid predictive model | MARS–NLRM | 0.78 | 0.053 | 0.034 | 0.71 | 0.061 | 0.039 | RF–NLRM | 0.86 | 0.042 | 0.028 | 0.70 | 0.062 | 0.039 | SVM–NLRM | 0.76 | 0.056 | 0.032 | 0.69 | 0.063 | 0.037 |
|
|