Hello, dear Knimers, I have another issue: I made a survey with several professionals, asking questions about their business situation over a series of months. I investigated three different variables (COVID-related rates) as potential determinants of bad performances of their businesses. These questions had six different alternatives as possible answers. These options are categorical ordinal variables, with string values, representing percentage intervals (e.g., <5%; 5-10%;… ; >75%). I have already converted them to other formats for labeling each value, both as string and integer values. Now, I wish to know if there is (and which are) the algorithms available in Knime for the desired multi-class classification. Thanks for any help. B.R., Rogério.

Help with multi-label classification

mlauber71 September 9, 2022, 3:37am 2

@rogerius1st you could check the links about multiclass in my machine learning collection

And then these sample workflows:

A good statistic to go for might be LogLoss:

1 Like

Optimisation of SVM parameters