Please consider citing the scikit-learn.
Mean square error regression loss
Positive floating point value: the best value is 0.0.
return the mean square error
y_true : array-like
y_pred : array-like
loss : float