Issue
This picture shows my understanding of the process of running a model and my question about the process of training a model.
Can anyone explain what happens then?
Thanks to @Michael Hearn 's answer, I grabbed information and filled the picture:
Solution
Hopefully someone says a better answer after mine later... So basically you are asking how a neural network functions. So the model is trained on the training data going in it then passes through nodes which run functions on the information and change it as it passes through. Then an output is generated. The output is compared to the test answer. Then through the not magical process of backpropagation the network is corrected. Nodes that were heavily responsible for giving wrong answers on the training set are changed so the next pass through they will be more likely to give the right answer.
Answered By - Michael Hearn
0 comments:
Post a Comment
Note: Only a member of this blog may post a comment.