Table 2 Experimental comparison between different non-evolutionary machine learning techniques for burned area prediction. Median of training error and test error calculated over 50 runs of each technique.
From: Predicting Burned Areas of Forest Fires: an Artificial Intelligence Approach
Method | Training | Test |
|---|---|---|
GS-GP | 12.0 | 12.9 |
SVM polynomial kernel (second degree) | 12.3 | 13.6 |
Random forests | 13.2 | 14.3 |
Radial basis function network | 18.5 | 19.3 |
Linear regression | 18.9 | 19.6 |
ST-GP | 13.8 | 21.0 |
Isotonic regression | 18.3 | 21.5 |
Neural networks | 20.3 | 33.8 |