BigCal Calibrations and Neural Networks

Found a bug in code used to train previous (gamma) NN. The results were biased towards higher energies. Previous statistics were low but now they have improved.

Parallel

Network inputs (before and after):

Energy networks (before and after):

X correction networks (before and after):

Y correction networks (before and after):

[ICO]NameLast modifiedSizeDescription

[PARENTDIR]Parent Directory   -  
[IMG]NNParaGammaXYCorrect..>2015-01-09 17:56 113K 
[IMG]NNParaGammaEXYCorrec..>2015-01-09 17:56 130K 
[IMG]NNParaGammaEXYCorrec..>2015-01-09 17:56 50K 
[IMG]NNParaGammaEXYCorrec..>2015-01-09 17:56 49K 
[IMG]NNParaGammaEXYCorrec..>2015-01-09 17:56 46K 
[IMG]NNParaGammaEXYCorrec..>2015-01-09 17:56 49K 
[IMG]NNParaGammaEXYCorrec..>2015-01-09 17:56 51K 
[IMG]NNParaGammaEXYCorrec..>2015-01-09 17:56 50K 
[IMG]NNParaElectronXYCorr..>2015-01-09 17:56 130K 

Apache/2.4.62 (Red Hat Enterprise Linux) Server at hallcweb.jlab.org Port 443