Values generated for the confusion matrix for a forest can cause the table to stretch for some columns, but not others, which can make interpreting the table more difficult, visually. Additionally, when classification is done along some dimensions (cylinders, for example), only numerical values are present. Various forests were generated during testing, and only one subject commented on the lack of labels on the matrix. For another subject, this connection was obvious. Aside: This issue may be solved by simply providing a transcript of the parameters used in the forest generation
Values generated for the confusion matrix for a forest can cause the table to stretch for some columns, but not others, which can make interpreting the table more difficult, visually. Additionally, when classification is done along some dimensions (cylinders, for example), only numerical values are present. Various forests were generated during testing, and only one subject commented on the lack of labels on the matrix. For another subject, this connection was obvious. Aside: This issue may be solved by simply providing a transcript of the parameters used in the forest generation
Severity Rating: 2