Back to brainteasers
Consider the data (x, y) drawn from a Gaussian distribution centered
at the origin with covariance matrix Σ = [1, −0.5; −0.5, 1]. Which data-
point is most likely to be considered an outlier?
Mark as solved