PSP benchmark

The ICOS PSP benchmarks repository contains an adjustable real-world family of benchmarks suitable for testing the scalability of classification/regression methods. When we test a machine learning method we usually choose a test suite containing datasets with a broad set of characteristics, as we are interested in knowing how the learning method reacts to a veriety of scenarios. The PSP field provides us with a whole family of real-world classification/regression problems that can be adjusted almost arbitrarily in terms of number of variables, number of classes, class balance, etc. Thus, these datasets are an ideal benchmark suite for data mining methods.

How to cite?

If you use datasets from our benchmark, please cite it as follows:

• DOI BibTeX
  M. Stout, J. Bacardit, J.D. Hirst, N. Krasnogor
  Prediction of recursive convex hull class assignments for protein residues
  in Bioinformatics, 24(7):916-923, April 2008

  @ARTICLE{Stout2008,
    title = {Prediction of recursive convex hull class assignments for protein residues},
    author = {Stout, M. and Bacardit, J. and Hirst, J. D. and Krasnogor, N.},
    year = 2008,
    doi = {10.1093/bioinformatics/btn050},
    month = apr,
    journal = {Bioinformatics},
    volume = {24},
    number = {7},
    pages = {916--923}
  }
  

@misc{PSPbenchmarks,
    author = "Jaume Bacardit and Natalio Krasnogor",
    title = "The ICOS PSP benchmarks repository",
    note = "(http://icos.cs.nott.ac.uk/datasets/psp_benchmark.html)",
    year = "2008"
}