
For high-dimensional data (D>1), only a few things change. Instead of estimating the mean and variance for each Gaussian, now we estimate the mean and the covariance. The covariance is a squared matrix of shape (D, D) — where D represents the data dimensionality. Be…