From: Leading indicators and maritime safety: predicting future risk with a machine learning approach
Modell | Â | Â | Score | Â | Â |
---|---|---|---|---|---|
 |  | MSE | r2 | Expl. Var. | |
DummyRegressor | Â | I | 0.3133 | Â | Â |
LinearRegression | Â | II | 0.3115 | 0.0005 | Â |
RF Model including PSC data | n_estimators = 1000, max_depth = 4 | III | 0.3016 | 0.0247 | 0.0355 |
 | Comp. to I | 96% |  |  | |
Opti. hyperparam. | Â | 0.3014 | 0.0311 | Â | |
 | Comp. to III | 100% | 126% |  | |
RF Model without PSC data | n_estimators = 1000, max_depth = 4 | IV | 0.3040 | 0.0202 | 0.0290 |
 | Comp. to I | 97% |  |  | |
 | Comp. to III | 101% | 82% | 82% | |
Opti. hyperparam. | Â | 0.3018 | 0.0287 | Â | |
 | Comp. to IV | 99% | 143% |  |