Calculates popular classifier performance values based on actual classes, and predicted values.
Different measures calculated for multi-class problems. Suffix _u refers to micro averaging, emphasizing single class values. Suffix _M refers to macro averaging, describing the overall performance.
See more information: Sokolova, M., Lapalme, G. (2009). A systematic analysis of performance measures for classification tasks. Information Processing and Management, 45, p. 427-437.
http://rali.iro.umontreal.ca/rali/sites/default/files/publis/SokolovaLapalme-JIPM09.pdf
Version | 1.0 |
---|---|
Bundle | tools |
Categories | Classification |
Authors | Ville Rantanen (ville.rantanen@helsinki.fi) |
Issue tracker | View/Report issues |
Requires | R |
Source files | component.xml ClassPerf.r |
Usage | Example with default values |
Name | Type | Mandatory | Description |
---|---|---|---|
in | CSV | Mandatory | Input data |
Name | Type | Description |
---|---|---|
out | CSV | A single row CSV containing the performance values. |
Name | Type | Default | Description |
---|---|---|---|
classCol | string | (no default) | Column name of the classes. |
conditions | string | "*" | Comma separated list of possible classes. If *, contents autodetected from the two columns. |
name | string | "Performance" | contents of the first column of the output |
predCol | string | (no default) | Name of the prediction column. |
Test case | Parameters▼ | IN in |
OUT out |
|||
---|---|---|---|---|---|---|
case1_binary | properties | in | out | |||
classCol=PFS, |
||||||
case1_binary_good | properties | in | out | |||
classCol=Group, |
||||||
case2_multi_missing_class | properties | in | out | |||
classCol=PFS, |
||||||
case3_multiclass | properties | in | out | |||
classCol=PFS, |
||||||
case4_binary_with_multidata | properties | in | out | |||
classCol=PFS, |