How to measure SP accuracy improvement after learning process


#1

Hello,
I am using the algorithm API, according to the example in link
I training only the SP, but my question is how I can know that the performance of the SP was improved during the learning process? (Any built-in matric ?)
Thanks


#2

If you send in similar input to the SP, it should produce similar output. The more different the input is the more different the output should be.