Feature selection is the process of selecting a subset of relevant features (variables, predictors) for use in model construction. PageRank (or PR in short) is a mathematical algorithm that evaluates the quality and quantity of links to a webpage. MRMD2.0 uses the idea of pagerank, combined with some other feature selection methods,including
  • MRMD
  • MRMR
  • MIC
  • etc.

MRMD2.0 can reduce the dimension of user-specified feature intervals.The supported file formats are :
  • arff
  • csv
  • libsvm
If you are not sure about the specific format of the data set,you can see it by clicking on the Dataset above.
We provide Web Server, you only need to provide the dataset to the WebServer to select the feature.Please click Help for specific usage.