Consider the Iris dataset, which has 4 attributes. KDE(kernel density estimation) is univariate analysis then how can we get overall kernel density estimation of the iris dataset


1 Answer 1


With multivariate data, you need to use multivariate kernel density estimation, that differs from univariate case by either using product of univariate kernels, or multivariate kernels (e.g. instead of univariate Gaussian, multivariate normal density as kernel).

  • $\begingroup$ That's surely right in principle. In practice, good luck visualizing a surface over all the variables based on just 150 observations for the Iris data. I would focus on univariate and bivariate density estimation at most. $\endgroup$
    – Nick Cox
    Commented Jun 30, 2020 at 15:14
  • $\begingroup$ @NickCox it doesn’t say that the point is plotting it, otherwise agree. $\endgroup$
    – Tim
    Commented Jun 30, 2020 at 15:24
  • 1
    $\begingroup$ A density estimate for any other purpose could be useful, but at the same time it would hard to know how far to trust it if you can't see it. I wouldn't trust a univariate estimate I couldn't see. $\endgroup$
    – Nick Cox
    Commented Jun 30, 2020 at 15:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.