Neural Pattern Recognition stop criteria

Daniel Grebler

2018 4 11

0 个回答

6 次查看（30 天）

0 个投票

Hey, I need to be able to stop the training in criteria of the confusion matrix.

My training always end up with extreme bad recognition (I cant change nor add more data) *image 1 yet I see during the training that at a brief point of time the matrix received a valid good result for both targets, yet until the training ends - it get back to final matrix.

could i enforce the training to stop at a given confusion matrix values (<of >of )?

Thank you in advance