The current description for the tag
feature-extraction is the following:
Variables (used for prediction or explication) used in regression or regression-like models (like clustering, discrimination). Use this tag for questions about constructing such variables or selecting the best among them.
However, feature extraction is a step that is not restricted to be use in regression-like models. It is only a data representation used as the first step in most data processing approaches.
The current tag description is misleading and questions that should use this tag are currently not using it if not related to regression models like (this question for example).
What are your suggestions? Should we change the description? and if yes for what?
My suggestion would be along the lines of:
In machine learning, pattern recognition and data modeling in general, feature extraction is the action of computing variables that contain the relevant information from the data, so that the desired task can be performed by using a reduced representation instead of the complete initial data.
(This is inspired from the Wikipedia page "Feature extraction")