Figure 2
From: A new strategy to map landslides with a generalized convolutional neural network
The spider plots (A–J) show the performance of CNNs trained with conventional learning (dashed line in pink) and combined learning (solid line in black) approaches. Every row has models evaluated for a specific study area at different base resolutions. The precision and recall scores are plotted on the X-axis, and the MCC and F1 scores are plotted on the Y-axis. (K) illustrates the shape expected from a good model which can be differentiated from models with lower precision or recall scores. The envelope of a better model will have a higher MCC/F1 Score and will enclose the plot from a weaker model. (Res. = Resolution).
