Hi,

Andrew J termed the cost function the squared error cost function.

I doubt it is not only squared, but he has done the average by multiplying with (1/2*m) in the formula.

So, I think it should be called a mean squared error function rather than calling squared error function.

TMosh
2
Thanks for your suggestion.

Hey TMosh, Iâ€™m not sure. Itâ€™s just my thought, So can we term it as a mean squared error?

TMosh
4
Yes, that seems fine.

I think Andrew was using â€śsquared errorâ€ť as a cue that itâ€™s based on the squares of the errors.

There are other cost functions with different bases that you will see very soon.

1 Like