In a Neural Network Regression Plot R means R^2?

Hello, I'm trying to run a neural network fitting regression and when I get the regression plot I can see R everywere instead of R square. Is it the same thing or do I have to make any adjustment to this value? Thank you

Answer 1

[Abhishek]

Hi Joan,

I guess you are trying to know what 'R' in the Neural Network Fitting plot signifies and is it different from 'R^2'.

To provide a measure of how ‘good’ the predictions of the model are compared to actual outcomes, coefficient of correlation or coefficient of determination could be used.

‘R’ in the plot is coefficient of correlation. To get coefficient of determination ‘R^2’, you will have to square the values. Also, if you hover over the Model Summary Table in the Neural Network Fitting App, you can get brief details about parameters like 'MSE' and 'R'.

For further reading, you can refer to following documentation pages: -

• Fit Data with a Shallow Neural Network - MATLAB & Simulink (mathworks.com)
• Coefficient of Determination (R-Squared) - MATLAB & Simulink (mathworks.com)