What is RMSE?

RMSE is the root-mean-square error, defined as sqrt(mean(e_1^2)) where e_1 is the list of residuals of the regression. The RMSE is used in cases where there is not a clear distinction between the independent and dependent variables in a model. Depending on the form of the model, there are 3 different statistics that might be shown: 

The Pearson correlation coefficient, r: appropriate when regressing a linear model with both slope and intercept, e.g. y_1 ~ m x_1 + b

The coefficient of determination, R^2: appropriate for non-linear models of the form y_1 ~ f(x_1, x_2, ...)

The root-mean-square error: appropriate for general models of the form f(y_1) ~ g(x_1), or h(x_1, y_1) ~ 0.

Have more questions? Submit a request


  • 3
    Стилиян Петров

    i still don't understand nothing!

  • 0
    Michael Chauvin

    What if it only gives you capital R^2, then how do you find lowercase r?

    Edited by Michael Chauvin
  • 0

    How do I change to least squares? It only does rmse when i type in y_1~mx_1+b

    Edited by Smcginley
  • 0
    Hassan Sheidaee

    You can use these free online calculators:
    It is so simple

  • 0
    Yacine Naji

    I still don't understand


Article is closed for comments.
Powered by Zendesk