0
$\begingroup$

Density-reachability is a canonical extension of direct density-reachability. What is meant by canonical extension?

Source: A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with Noise https://www.aaai.org/Papers/KDD/1996/KDD96-037.pdf

$\endgroup$
  • 3
    $\begingroup$ Could you give the source of the quote (with hyperlink if possible)? $\endgroup$ – Peter Taylor May 1 '18 at 11:15
  • $\begingroup$ @PeterTaylor I added the title and a direct link to the paper where it was used $\endgroup$ – Vincent May 2 '18 at 11:19
2
$\begingroup$

The term 'extension' could generally be interpreted in mathematical sense1 as an enlargement of the domain of a pre-existing function.

In your case, assume Density-reachability could be described as a process $d\colon A \rightarrow B$ mapping inputs from domain $A$ to output in domain $B$. An enlargement of domain could be viewed as another map $e\colon A \rightarrow A'$ which maps where the old inputs in $A$ could be thought of as special cases in subset of $A'$ that is the image of map $e$.

An extension then could be described as any map $x\colon A' \rightarrow B$ which statisfy $$x \circ e = d$$ There could be many such $x$, so picking extension is a choice. The above equation basically expresses that

If I pick $a \in A$ from the old domain, translate it to the new domain as $e(a)$, then use this new tool $x$ which could be thought of as an extension of old tool $d$ along the translation $e$, then the final result must agree with just using the old tool $d$ to begin with.

or $$x \circ e(a) = d(a)\qquad; \forall a \in A$$

The term 'canonical' simply implies that the choice of extension is the obvious, natural one of the situation.

1 At least from the perspective of category theory.

$\endgroup$

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.