Do predictions when I

mariaj

I have applied a pretty sucky model using linear regression and cross-validation, but still need to understand this. My prediction for example was 1000. 

root_mean_squared_error: 952.712 +/- 111.892 

absolute_error: 645.593 +/- 34.608 (micro average: 645.593 +/- 708.616)

Do I have to adjust my predictions or do the models predictions already include an average error? 

Thanks!!

MartinLiebig

Hi,

this means: If you predict 2000, then with a 68% propability the real value is in 2000+- 645.

Does this make more sense?

Best,

Martin

MartinLiebig

Hi,

this means: If you predict 2000, then with a 68% propability the real value is in 2000+- 645.

Does this make more sense?

Best,

Martin

mariaj

Thanks!