Page 1 of 1

Aforge BackPropagationLearning multiple i/o datasets

PostPosted: Sun Sep 01, 2019 2:16 am
by razraz
Hi, I have a question. I am new to ML and Forge for that matter. I am curious about using BackPropagationLearning() with multiple datasets.
I have a 100 data sets. 100 unique input data sets and 100 sets of corresponding output data for each input set.
Each input and outputs data set consists of 3d vectors.

How do I train with a 100 sets of data? do I run BackPropagationLearning() 100 times using each of the 100 data sets and then average the results?

accord-framework.net/docs/html/T_Accord_Neuro_Learning_BackPropagationLearning.htm RunEpoch

Thanks in advance