0
$\begingroup$

This is a question about proper terminology related with what is understood with "Semi-Supervised Classification".

This is my context:

  • I have a rule-based classifier.
  • I know for sure I can classify my data in a number of N classes.
  • I have also a "small" dictionary where my rules check (during classification process) for every observation iterated.

Is this a semi-supervised classification?

For extra points: If the dictionary was built by-hand from a curation process, is this dictionary considered a "training set"?

$\endgroup$
0
$\begingroup$

The terms supervised classification and semi-supervised classification refer to classification based on machine learning methods. Machine learning models are trained with labeled data or partially labeled data. Supervised and semi-supervised learning algorithms analyze training data and produces inferred functions.

And rule-based classification are considered to belong to unsupervised classification. No labeled are needed to infer the classification rules mathematically. The "small" dictionary does not provided formal labeled data, and it is also part of the rule-based classification model.

$\endgroup$

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.