classification problem

[Overflow10]

Hi, I have a question.
I want to train a classifier that is able to detect a single specified speech word. I train the classifier with a training set that consists of positive samples: the features of the specified word. My problem is the training set should contain the against samples too, but they are very lots samples! How can I train a classifier with an infinity of against samples?
Thanks.

 

[harsha_jois]

Hi overflow10.
You do not need to train for negative samples. In your condition for negation, just provide the word, syllables, speech patterns with != (not equal to) condition. That would suffice.
If you're using a tool like tanagra, it will be easier.
Hop this helps.