I try to use Support Vector Machine (SVM) in Statistical [one factor] analysis to highlight important features, but the SVM in MA yields features just by order, without any sorting.
I tried to use example data (1) human_cachexia.csv (Data Filtering is not asked, autoscale), and this works well, the features are ranked.
If to use example data (6) lcms_table.csv (Data Filtering - None, autoscale), this again yields no sorting. Please see picture attached. I suppose there is some bug here.
This just means that all these features have equal chances to be selected by the model. No features are more important than others. This happens when many features are redundant, such as in untargeted metabolomics data