I want to perform attribute selection in Weka, but my dataset is rather big, and the program runs quite a while. That's why I want to see the current best set of attributes found. How do I do it? For example, genetic search has the "Report Frequency" parameter, but all the results are shown after the whole search is finished, that's not what I need.
There is no progress bar, so I don't even know for how long will I have to wait...
Feature or Attribute selection is a standard problem in data-mining and Machine learning domains.
If you want to select a good set of attributes, you must preprocess your data by ranking attributes based on their quality. Ranking Methods such as p-metric or t-statistic are popular, which are based on statistical measures. One cannot simply go about by randomly selecting attributes from a large set without any sort of intuition on the nature of attributes.
