hard20 pts · first answer countsregression-metrics

Negative R-squared on test data

Your regression model reports R^2 = -0.3 on the test set. What does this mean?

A.R^2 can never be negative, so the metric code is brokenB.It predicts worse than always guessing the test-set meanC.The features are perfectly correlated with each otherD.The model underfits, so R^2 was clipped below zero

Discussion 0

Join free to post in the discussion.

Negative R-squared on test data

Discussion 0

Related questions