Hello,
Woud you mind if explain for me the function to evaluate a context:
outsums[oid] = Math.exp((outsums[oid]*fval) + ((1.0 -(numfeats[oid]/correctionConstant))* correctionParam));
I dont understand when relate to the loglinear computing probability function p(x,c) in theory of Maximum Entropy.
I'm sorry if this question in trivial for you, because I 'm a a new beginer in Maximum Entropy model and MaxEnt package.
Thank you!
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
hi,
This is the line where the sums of logs are converted out of the log space for normalizing. That's the exp anyway. The other part is adding the correction constant and its parameter value to the existing sum.
Hope this helps...Tom
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello,
Woud you mind if explain for me the function to evaluate a context:
outsums[oid] = Math.exp((outsums[oid]*fval) + ((1.0 -(numfeats[oid]/correctionConstant))* correctionParam));
I dont understand when relate to the loglinear computing probability function p(x,c) in theory of Maximum Entropy.
I'm sorry if this question in trivial for you, because I 'm a a new beginer in Maximum Entropy model and MaxEnt package.
Thank you!
hi,
This is the line where the sums of logs are converted out of the log space for normalizing. That's the exp anyway. The other part is adding the correction constant and its parameter value to the existing sum.
Hope this helps...Tom