1
$\begingroup$

I am doing some testing in recommender systems with extended epinions dataset, I implemented two models, model A give me RMSE of 0.5387 and MAE of 0.3111 and model B gave me RMSE of 0.5121 and MAE of 0.3244.

I really don't know how should I interpret this degrade in MAE error, Can anyone help me about this?

$\endgroup$
2
$\begingroup$

RMSE leads to estimating mean and is more sensitive to extreme cases and outliers. MAE leads to estimating median and is more robust to outliers. If RMSE dropped but MAE increased, it means that your model is now better at accounting for extreme cases, but the solution may be less robust.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.