how can i plot R2(regression line) between target and prediction

Please explain how can i plot R2 as regression line between target and prediction in matlab

Can some one please share the code ?

i am attaching a sample graph.

Answer 1

[Divija Aleti]

Hi,

Firstly, as R2 is a scalar, it cannot be plotted.

To plot the regression line, follow the example shown in the link below:

plotregression

However, to get a similar output as shown in the above graph, you will have to make changes in the example. Refer to the sample code given below which is an extension of the example given in the link.

[x,t] = simplefit_dataset;
net = feedforwardnet(10);
net = train(net,x,t);
y = net(x);
plotregression(y,t,'Regression') % Reversed the order of t and y, as we want Target on the y-axis and Prediction on the x-axis
% Customization of title, x-label and y-label
a=gca;
title(a,'Test Data, R^2=0.9893')
xlabel(a,'Prediction')
ylabel(a,'Target')

Output:

For calculating the value of R-squared(R2), which is given as the title of the graph, have a look at the following links:

Linear Regression

Coefficient of Determination (R-Squared)

Hope this helps!

Regards,

Divija

Answer 2

[pathakunta]

Firstly, as R2 is a scalar, it cannot be plotted. To plot the regression line, follow the example shown in the link below: plotregression However, to get a similar output as shown in the above graph, you will have to make changes in the example. Refer to the sample code given below which is an extension of the example given in the link. [x,t] = simplefit_dataset; net = feedforwardnet(10); net = train(net,x,t); y = net(x); plotregression(y,t,'Regression') % Reversed the order of t and y, as we want Target on the y-axis and Prediction on the x-axis % Customization of title, x-label and y-label a=gca; title(a,'Test Data, R^2=0.9893') xlabel(a,'Prediction') ylabel(a,'Target') Output:

For calculating the value of R-squared(R2), which is given as the title of the graph, have a look at the following links: Linear Regression Coefficient of Determination (R-Squared) Hope this helps!