3 ms·
Totally agree I suspect the bias currently giving me P(detected|negative)>P(detected|positive) is resulting from my simplification to looking only at 1-grams
by markovbling 12y ago
Totally agree
I suspect the bias currently giving me P(detected|negative)>P(detected|positive) is resulting from my simplification to looking only at 1-grams