Auto Model Regression Results on Absolute Error and Squared Error

Nayli
Nayli New Altair Community Member
edited November 5 in Community Q&A
I have inquiry on absolute error and squared error. Does this same with mean absolute error (MAE) and mean squared error (MSE)? I did read the documentation but I'm still slightly confuse.

Another question is does the relative error is same as important for regression?
Would be great if someone can explain these to me.

Thank you
Tagged:

Best Answer

  • Roland Jones_21245
    Roland Jones_21245
    Altair Employee
    Answer ✓
    Hi @Nayli,

    The MAE is the mean error for each of the datapoints in the test dataset. MSE is the mean of the square of these errors.

    Relative error is the average of the deviations divided by the actual value. I believe this is the formula: (1 / N) Σ_n (|Y_n - f(X_n)| / |Y_n|)

    Best,

    Roland

Answers

  • Roland Jones_21245
    Roland Jones_21245
    Altair Employee
    Answer ✓
    Hi @Nayli,

    The MAE is the mean error for each of the datapoints in the test dataset. MSE is the mean of the square of these errors.

    Relative error is the average of the deviations divided by the actual value. I believe this is the formula: (1 / N) Σ_n (|Y_n - f(X_n)| / |Y_n|)

    Best,

    Roland

Categories