Equations · Machine learning
Mean squared error
The average of the squared differences between predictions and true values.
MSE = (1/n) ∑ᵢ (yᵢ − ŷᵢ)²
Aethel says
M S E equals the quantity 1 over n, times the sum over i of the quantity y sub i minus y hat sub i, squared
What it says
For each example, take the error, square it so that misses in either direction count, and average over all n examples. Large errors weigh much more than small ones.
Minimising it gives ordinary least squares in linear regression.
Symbol by symbol
| Symbol | Say it | Means |
|---|---|---|
| MSE | M S E | the mean squared error |
| yᵢ | y sub i | the true value for example i |
| ŷᵢ | y hat sub i | the prediction for example i |
| n | n | the number of examples |
Where you meet it
Regression, forecasting and the training of neural networks.
Questions
How do you say Mean squared error out loud?
MSE = (1/n) ∑ᵢ (yᵢ − ŷᵢ)² is read: “M S E equals the quantity 1 over n, times the sum over i of the quantity y sub i minus y hat sub i, squared.” That is exactly what Aethel says when it meets the formula in a book.
What does MSE mean in Mean squared error?
MSE (said “M S E”) is the mean squared error.
What does yᵢ mean in Mean squared error?
yᵢ (said “y sub i”) is the true value for example i.
Can I listen to the equations in my own textbook like this?
Yes. Upload a PDF, EPUB or text file and Aethel reads it aloud with every equation spoken as maths, the way it reads this one, and highlights each sentence as it goes. You can start free.