Any recommendation on Hypothesis scoring limits? – Politepix /forums/topic/any-recommendation-on-hypothesis-scoring-limits/feed/ Tue, 23 Apr 2024 14:48:16 +0000 https://bbpress.org/?v=2.6.9 en-US /forums/topic/any-recommendation-on-hypothesis-scoring-limits/#post-1015440 <![CDATA[Any recommendation on Hypothesis scoring limits?]]> /forums/topic/any-recommendation-on-hypothesis-scoring-limits/#post-1015440 Tue, 22 Jan 2013 13:52:57 +0000 Geri Is there any recommendation on hypothesis scoring limits?
I’d need around 4 categories, I’m just prototypeing now.

You said “WORD”.
Seems you said “WORD”.
Probably you said something like “WORD”.
Sorry, I couldn’t understand what you said.

What could be the hypothesis scrore limits for these?

]]> /forums/topic/any-recommendation-on-hypothesis-scoring-limits/#post-1015441 <![CDATA[Reply To: Any recommendation on Hypothesis scoring limits?]]> /forums/topic/any-recommendation-on-hypothesis-scoring-limits/#post-1015441 Tue, 22 Jan 2013 14:08:29 +0000 Halle Winkler Hello,

Here are the previous discussions:

/forums/topic/how-does-recognition-score-works/
/forums/topic/scores-used-in-openears/

You can’t use ranges to limit because you don’t know what the environmental factors are, but something you can experiment with is detecting cases where the top n-best and the second-closest n-best are close and the overall range is large. i.e. 1-best is -50000 and 2-best is -50500 suggests to me that it’s a close call. But you can’t say “below -50000 is wrong”.

]]>