[mlpack-git] (blog) master: Remove comment. (4678f6a)

[gitdub at mlpack.org]

Repository : https://github.com/mlpack/blog
On branch  : master
Link       : https://github.com/mlpack/blog/compare/a3d44bdf7731169a5587918fd41414e29a5f0901...4678f6acd72427988c28b3275f8a67ef28c464e8

>---------------------------------------------------------------

commit 4678f6acd72427988c28b3275f8a67ef28c464e8
Author: MarcosPividori <marcos.pividori at gmail.com>
Date:   Mon Jul 18 15:55:45 2016 -0300

    Remove comment.


>---------------------------------------------------------------

4678f6acd72427988c28b3275f8a67ef28c464e8
 content/blog/MarcosWeekSeven.md | 4 ----
 1 file changed, 4 deletions(-)

diff --git a/content/blog/MarcosWeekSeven.md b/content/blog/MarcosWeekSeven.md
index 71929f4..069663b 100644
--- a/content/blog/MarcosWeekSeven.md
+++ b/content/blog/MarcosWeekSeven.md
@@ -18,9 +18,5 @@ As mentioned in last blog post, Hybrid Spill trees have both overlapping and non
 
 We can control the hybrid by varying $\tau$. If $\tau$ is zero, we have a pure spill tree with defeatist search, very efficient but not accurate enough. If $\tau$ is a very large number, then every node is a non-overlapping node and we get back to the traditional metric tree, with prunning rules, perfectly accurate but not very efficient. By setting different values for $\tau$, we have a trade-off between efficiency and accuracy.
 
-Then, which is the smallest value of $\tau$ that provide exact results?
-If we are sure that $\tau$ is a valid upper bound of the kth nearest neighbor of the query point $p_q$, this means: $dist(p_q, kth-n(p_q)) < \tau$, then we can be sure that we will get an exact solution.
-Of course, we can use smaller values of tau if we want an approximation.
-
 Next week, I will continue improving the implementation of Spill Trees and the Defeatist Seach and write some tests. Also, I have to continue thinking about possible alternatives to implement Hybrid SP-Tree Search as a dual tree algorithm. Follow the progress in: [[2]](https://github.com/MarcosPividori/mlpack/tree/spill-trees/src/mlpack/core/tree/spill_tree).