Confusion Matrix

Summary

Doc_BoxAlgorithm_ConfusionMatrix.png
  • Plugin name : Confusion Matrix
  • Version : 1.0
  • Author : Laurent Bonnet
  • Company : INRIA/IRISA
  • Short description : Make a confusion matrix out of classification results coming from one classifier.
  • Documentation template generation date : Oct 29 2020

Description

The Confusion Matrix box performs real-time computation of the confusion matrix of a given classifier. Confusion matrix can be used to measure the performance of a classifier. The confusion matrix output can be filled with either percentages or values. Optional colum and row can be added to give the sums of each row and column. The confusion matrix output can be displayed usig a BoxAlgorithm_MatrixDisplay.

Inputs

1. Targets

The stimulations that comes from the instruction flow, i.e. the targets that the classifier aims at.

  • Type identifier : Stimulations (0x6f752dd0, 0x082a321e)

2. Classification results

The classification results coming from a classifier. These stimulations will be compared to teh target to perform the computation.

  • Type identifier : Stimulations (0x6f752dd0, 0x082a321e)

Outputs

1. Confusion Matrix

The Confusion matrix.

  • Type identifier : Streamed matrix (0x544a003e, 0x6dcba5f6)

Settings

More box settings can be added for a multi class classifier. The default configuration uses 2 classes.

1. Percentages

Tells the box to put percentages or values in the matrix.

  • Type identifier : Boolean (0x2cdb2f0b, 0x12f231ea)
  • Default value : [ true ]

2. Sums

If checked, this option adds one row and one column that gives the sums of each row and column in the matrix.

  • Type identifier : Boolean (0x2cdb2f0b, 0x12f231ea)
  • Default value : [ false ]

3. Class 1

The stimulation label for the first class.

  • Type identifier : Stimulation (0x2c132d6e, 0x44ab0d97)
  • Default value : [ OVTK_StimulationId_Label_00 ]

4. Class 2

The stimulation label for the second class.

  • Type identifier : Stimulation (0x2c132d6e, 0x44ab0d97)
  • Default value : [ OVTK_StimulationId_Label_01 ]

Examples

Miscellaneous