Hi benglish, Now I get what you are meaning, but sorry I don't know why.
Anyway, I'm glad it seems that you finally did train and test on Lucene index. Koji -- http://soleami.com/blog/comparing-document-classification-functions-of-lucene-and-mahout.html (2014/06/27 21:02), benglish wrote:
Suppose I have the following code: 1. buildIndex(); 2. trainBaysianNetwork(); 3. testNetwork(); When I run just the first 2 lines (not the 3rd line) for more than 1 time, and then run the 3rd line, I get a better accuracy. The more I run just the first 2 lines, the better accuracy I gain. I want to know why it is in this way?! -- View this message in context: http://lucene.472066.n3.nabble.com/Train-Lucene-with-topic-defined-files-tp4141979p4144442.html Sent from the Lucene - General mailing list archive at Nabble.com.
