Fig 1.
Silhouette coefficient vs. number of clusters.
Fig 2.
Distortion score vs. number of clusters.
Fig 3.
SHAP feature importance and impact plot – cluster 1.
Fig 4.
SHAP feature importance and impact plot – cluster 2.
Fig 5.
SHAP feature importance and impact plot – cluster 3.
Fig 6.
SHAP feature importance and impact plot – cluster 4.
Fig 7.
SHAP feature importance and impact plot – cluster 5.
Fig 8.
Geographic distribution of analysed stock exchanges.
The map was created by the authors using Python’s Geopandas library and public domain data from Natural Earth (http://www.naturalearthdata.com). No copyright restrictions apply.
Table 1.
Summary of cluster characteristics and notable exchanges.