# Machine-learning (python) non-parametric continuous variables and output

There are various machine-learning techniques available, of which I know there is the (K) NN -> nearest neighbour.

However, it seems most non-parametric ML techniques need the input and output to be 'digitized'. Is it possible to have non-parametric ML techniques on continuous variables, both in input and output?

I know it is possible for linear (and non-linear) forms for parametric via regression, for continuous input/output.

Kind regards

• What do you mean by "digitized" and what do you mean by continuous? It's usually understood that a continuous random variable can take any value in $\mathbb{R}$. If this is what you mean, Python, R, MATLAB, Stata, Eviews, etc. all can do what you're looking for without any effort on your part. – Stéphane Apr 19 at 20:27