Compute Friedman and Nemenyi Statistical Tests for Multilabel Classification

This source code has the following R scripts

1. libraries.r
2. utils.r
3. ranking.R
4. Friedman-Nemenyi-v1.R
5. Friedman-Nemenyi-v2.R
6. radar_plots.R
7. example_fn.R
8. example_radar.R

The ranking.R generates the rankings for each of the multilabel measures according to their best value as follow:

The Friedman-Nemenyi_v1.R uses the tsutils package, while Friedman-Nemenyi_v2.R uses the scmamp package.

This source code does not provide the installation of any of the packages used, therefore, you must install the libraries used. Check libraries.R.

The scmamp package can give some installation headaches. For newer versions of R it is not available, so it is necessary to force the installation manually. Consult here and [here] (http://cran.nexr.com/web/packages/scmamp/index.html) all necessary dependencies and I suggest you to install via tar.gz.

For this code to work correctly, the files must be provided in CSV format. This code was developed to work with the 22 multi-label assessment measures. You can provide just one CSV file, or all 22 files (one for each measure). The CSV file format of the assessment measure should be as follows:

Along with this code are provided examples of csv files in the DATA folder

You can generate radar plots for your data. An example is provided in example_radar.R

You can see an example of how to use this code by referring to the example_fn.R.

This study is financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) - Finance Code 001

| Post-Graduate Program in Computer Science | Computer Department | Biomal | CAPES | Embarcados | Read Prensa | Linkedin Company | Linkedin Profile | Instagram | Facebook | Twitter | Twitch | Youtube |

Please contact me: [email protected]