$\begingroup$

I am using example from here:

Root Mean Squared Error:

1. Find out the difference between original and predicted values for each row.
2. Square the differences
3. Sum all squared differences
4. Take the average of the above sum
5. Take the square root of above average

Table:

We see from table that RMSE = 14.2.

Here is what and how I calculated from data presented in table and method described above table:

$(25-20)^{2} + (30-31)^{2} + (29-33)^{2} + (40-38)^{2} = 46$

$\frac{46}{4} = 11.5$

$\sqrt{11.5} = 3.39116$

So in table we have RMSE = 14.2 and from my calculation RMSE = 3.39116. Is there mistake in table or in my calculations?

$\endgroup$

1 Answer

$\begingroup$

Good day,

• You sure it is RMSE (Root Mean Square Error)?

• And not MAPE (Mean Absolute pErcentage Error?)

The page seems to be wrong, or your screenshot?

---

Actually someone was also confused by this.

---

"Hi FARUKH, I find the calculation of RMSE is wrong. Instead of taking the square of APE, it should be the square of the difference between the original and predicted value."

---

From the website at

Q. What is RMSE?

Root Mean Squared Error
1. Find out the difference between original and predicted values for each row.
2. Square the differences
3. Sum all squared differences
4. Take the average of the above sum
5. Take the square root of above average

• I also see another problem; with the number of columns. I Think it should be something on those lines:

On the webpage It's 5 columns. In your screenshot It's only 4 columns.

Your screenshot:

Original Predicted
20 25
31 30
33 29
38 40

on the webpage it looks slightly different, and is instead:

Original Predicted
10 12
14 13
18 15
20 23
11 15

---

(12-10)²+(14-13)²+(18-15)²+(20-23)²+(11-15)² = 39

39 (as in the webpage screenshot

@ SUM: 39

Mean(SUM): 7.8 39/5 = 7.8

RMSE: 2.79 $\sqrt 7,8$ = 2.7928

---

webpage screenshot:

This was a interesting question, I'll follow this.

Hope it solves itself.

Regards Will

$\endgroup$

Your Answer

Sign up or log in

Sign up using Google Sign up using Facebook Sign up using Email and Password Submit

Post as a guest

Post Your Answer Discard

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