Search code examples
algorithmclassificationweka

What is "parameters" in classification algorithm


If I run a 5-fold cross validation on a particular data set on Weka, I will have a set of optimized parameter.

What is actually this parameter result that I am getting from a 5-fold cross validation?

How can I use this parameter to run another 10-fold cross validation on weka?


Solution

  • Which classification algorithm are you using?

    Regardless, you can use weka's meta classifier, CVparameterSelection, to figure out the optimal values/combinations of input parameters for your classifier and then re-run with those values specified. For a pretty good explaination, check here: http://weka.wikispaces.com/Optimizing+parameters