Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization. It only takes a minute to sign up.
Sign up to join this community
I was calculating the root mean squared error for a regressive model, and while calculating, its mean squared error (MSE) came out to be less than 1 (>0), and its root (the root mean squared error, RMSE) as obvious came out to be greater than it. What does that mean?
like MSE = 0.0929285270191
and hence, RMSE = 0.894951677992
How do I see it?
Was there any problem in calculation or training of the model?
Is it that MSE is coming out very less that is not likely possible?
Your calculation must be wrong somewhere.
As the name says the RMSE is the square root of the MSE, so for MSE = 0.09 the RMSE would be 0.3. Yours is approximately off by a factor of 3.
It means that the $MSE$ is less than $1$, which might be spectacular performance, dreadful performance, or somewhere in between, depending on your problem. Performance out of context is fairly meaningless.
That's just how the square root works: $ x\in(0, 1)\implies x < \sqrt{x} $.
Note that the example values you gave are not possible, as the $RMSE$ corresponding to your $MSE = 0.0929285270191$ would be $RMSE=\sqrt{0.0929285270191}\approx 0.305$, though I think you just picked two numbers arbitrarily that satisfy $MSE < RMSE$.
