Reversed Predicted Categories?

Shubha

Hi,

I am learning an SVM model (hyperhyper) on a training exampleset and applying the model on the same exampleset(without the label).

My problem is this:

All the labels which were negative in the training data is positive in the predictions and all the labels which were positive in the training data is now negative in the predictions test data. And this is happening with the confidence of 1 for positive/negative.

Why is this happening with so much of confidence, when my  training and test data are the same? Am I missing some options to click?

Thanks,
Shubha

haddock

Shubha, there is absolutely no point in clogging up this forum with your ridiculous questions if you are not prepared to read or cannot understand the replies. This topic has been flogged to death recently http://rapid-i.com/rapidforum/index.php/topic,776.0.html

You have now posted 60 times in the last month, so are you being paid to do it, in which case we know where to send the bill, or do you do it for fun, in which case I would recommend some urgent therapy?