[mlpack-git] [mlpack/mlpack] Vantage point tree (#708)

lozhnikov notifications at github.com
Fri Jul 22 04:39:04 EDT 2016


> But note that we never visit (N_q0, N_r1) or (N_q0, N_r2), so base cases will not be performed between any points in N_q0 and N_r1 or N_r2, and the results end up being wrong.

That was my mistake, I did not point out the difference in the traversers. I use the single tree traverser for a query point and a reference node. But if a reference node contains a point I fire off all base cases `BaseCase(queryNode.Descendant(i), referenceNode.Point(0))` since we have not got the score algorithm for a query node and a reference point. Is it possible to implement the score algorithm for a query node and a reference point?

> Maybe there is an error somewhere, where the bounds are way looser than they need to be?

Could that happen due to a huge overlap?

---
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/mlpack/mlpack/pull/708#issuecomment-234487133
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.cc.gatech.edu/pipermail/mlpack-git/attachments/20160722/7ec33563/attachment.html>


More information about the mlpack-git mailing list