Which unit of measurement has the accuracy, mean absolut error and mean squared error of the sklearn metrics? and are they used for regression problems? For example i have a regressor.score(X_test,y_test)) of about 0.99469. So the tested Model is 0.99469 in comparison to the real data? or do you say percent of 99%? The same with the MAE and MSE.
The other question is, i red about the confusion matrix for classification problems and the accuracy is the value calculated out of the confusion matrix. I use a multiple linear regression, so can i use these metrics for my regression?
To answer your first question, the metric such as accuracy can't be used for regression problems. And yes, you are right. The accuracy is calculated using the confusion matrix, but since you have a regression problem, you can't get the confusion matrix as it is an output of the classification problem but you have a regression problem in place.
Moreover, the right metrics for the regression problem are mean squared error, mean absolute error and the R squared value. By default the regressor.score(X_test,y_test)) function gives out the R squared value.
To make it easy, the closer the value is to 1 (0.99469 in your case), the better is your model. And it seems like your model is performing really well.