correlation is very low,i need solution how to increase r2

after running partial least square regression for my project work i'm getting r2 as 0.004, how can i increase the r2 to 0.8.please give me a solution

thanking you

cdy

[Merged from duplicate question]

really i'm fed off please anyone give me solution and help me, after plotting measured Vs predicted scatter plot i'm getting r2 as 0.004, please help me how can i increase it to 0.9. i have done all type of experiments but dint get solution.

[José-Luis]

Do you realize the near-impossibility to answer your question considering the details you have given? To get better R2, you can modify your data, but that's generally frowned upon, or you can modify your model. The latter alternative is preferred, but it is pretty much impossible to help you without knowing what you are actually trying to do. People in this forum are willing to help, but forcing us to guess is a waste of everybody's time.

[chaya d y]

i'm actually trying to correlate the organic crabon values of laboratory experiment values and using these organic carbon and reflectance value i'm going to run partial least square regression from PLSR i'll predict organic carbon using those predicted and measured values i'm going to find r2 value, i'm getting r2 as 0.004 but i want to increase it to 0.9.plz tell me how can i modify my model or any other solution..

thanking you cdy

[Walter Roberson]

Perhaps PLSR is not suitable for your purposes. Why not show us your scatter plot?

http://www.mathworks.com/matlabcentral/answers/7924-where-can-i-upload-images-and-files-for-use-on-matlab-answers