What does regression value in neural networks denote

3 views (last 30 days)
Elysi Cochin
Elysi Cochin on 30 Jun 2021
Commented: KSSV on 2 Jul 2021
What does the R value of Training, Validation, Testing and All denote, in the below plots
When i compute accuracy using classperf, i get another value such as 0.97307
So, how is regression value related to accuracy or is it the same?
Also what does the xlabel values in the below graph mean?

Accepted Answer

KSSV on 1 Jul 2021
Regression shows you how close the two datasets are. Regresion lies between [0,1]. The more the value the close the two datasets (in phase) are.
There is a formula to calculate the regression. Simple google will show you up the formula.
KSSV on 2 Jul 2021
Okay...So the basics of NN is you have input-x and output-y; in NN you try to get a function y = f(x). Here f are the weights. NN gets weights by learnng the given input, output. When weights are generated, it will calculate it's output and comapres with the given output to adjust weights. So here regression is inbetween the NN generated output and the output provided by the user.

Sign in to comment.

More Answers (0)


Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!