Fig. 11: SHAP global importance of cost factors and distribution of sample-level effects.
From: Machine-learning-based reconstruction of Ming-dynasty defensive corridors in Yuxian
a It is a SHAP-based beeswarm plot of cost factors. Each point represents the SHAP value of a single sample; the horizontal axis denotes the positive and negative contributions to the model output, the vertical position corresponds to different factors, and the color of the points indicates the magnitude of factor values. This plot is used to depict the influence direction and dispersion degree of each factor at the sample level; b shows a bar chart of global importance and a percentage pie chart, which display the proportion of the mean absolute SHAP value of each factor in the total explanatory power.
