I finished the course 1 but one question that didn’t find the answer for was how do I choose the right variable for linear regression. i.e how do I measure the effectivness of linear regression ( for example does the model predict values at 90% accuracy etc). I may have missed it. Can anyone guide me?
For linear regression, use metrics like
mean squared error or
mean absolute error on the test set to get a picture of how well your model generalizes to unseen data.
Accuracy is a metric applicable to logistic regression but the same idea holds true. Measure model performance on the test set.