4
$\begingroup$

I am studying decision trees and machine learning. I have encountered with this word several time. Although, it seems that it not an important concept, I cannot understand this word in machine learning literature. For example:

if the test does not significantly satisfy the criteria, then it creates a leaf and computes the prototype of the instances belonging to the leaf" .

what does prototype in this sentence mean?

According to Wikipedia:

A prototype is an early sample, model, or release of a product built to test a concept or process or to act as a thing to be replicated or learned from.

But it does not make sense.

$\endgroup$
  • $\begingroup$ I have updated my answer with an addition of the technique on k-prototypes $\endgroup$ – Laurent Duval Mar 13 '18 at 9:41
2
$\begingroup$

A prototype is an element of the data space that represents a group of elements. On the context of clustering (e.g. under a leaf), a cluster prototype serves to characterize the cluster, their elements. You can have a look at Cluster analysis: basic concepts and algorithms for instance, taken from Introduction to data mining.

In a more specific acceptation, it denotes an instance of clustering for combined numerical and categorical values, in the spirit of k-means, k-modes, k-medians, see for instance Into the world of clustering algorithms: k-means, k-modes and k-prototypes:

For numerical and categorical data, another extension of these algorithms exists, basically combining k-means and k-modes. It is called k-prototypes.

$\endgroup$
  • $\begingroup$ the link does not work. could you correct it? $\endgroup$ – David Apr 4 '16 at 16:42
  • 1
    $\begingroup$ Link is corrected $\endgroup$ – Laurent Duval Apr 4 '16 at 16:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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