<p>Continuing with my previous comment:</p>

<p>I think we have 3 independent decisions to make, when building a binary space tree:</p>

<ul>
<li>
<strong>Which kind of splitting Hyperplane:</strong> <em>Axis-Orthogonal</em>, or <em>General</em> (not necessarily Axis-Orthogonal).</li>
<li>
<strong>How to choose the hyperplane:</strong> <em>Random</em> or <em>Widest range</em> (the one with widest range of projection values).</li>
<li>
<strong>Which Splitting value we wil consider:</strong> <em>MidPoint value</em>, <em>Mean value</em>, <em>Median Value</em>.</li>
</ul>

<p>Every combination is valid, for example:</p>

<ul>
<li>
<em>Axis-Orthogonal - Random - MidPoint</em>.</li>
<li>
<em>General - Widest range - Mean Value</em>.</li>
<li>
<em>Axis-Orthogonal - Widest range - Median Value</em>.</li>
<li>etc.</li>
</ul>

<p>Now, for Spill Trees, I conside all the combinations between Hyperplane type and Splitting Value, as 2 template parameters. We could add a third template parameter that represents the criterion to choose the hyperplane, between Random and Widest range.</p>

<p style="font-size:small;-webkit-text-size-adjust:none;color:#666;">&mdash;<br />You are receiving this because you are subscribed to this thread.<br />Reply to this email directly, <a href="https://github.com/mlpack/mlpack/pull/726#issuecomment-237876657">view it on GitHub</a>, or <a href="https://github.com/notifications/unsubscribe-auth/AJ4bFH0QPX7P9y95xjm8N6b_DBiTbFxFks5qc1KPgaJpZM4JOuGE">mute the thread</a>.<img alt="" height="1" src="https://github.com/notifications/beacon/AJ4bFIOEoQBMWoq7MvGPQjrH8MdjJekmks5qc1KPgaJpZM4JOuGE.gif" width="1" /></p>
<div itemscope itemtype="http://schema.org/EmailMessage">
<div itemprop="action" itemscope itemtype="http://schema.org/ViewAction">
  <link itemprop="url" href="https://github.com/mlpack/mlpack/pull/726#issuecomment-237876657"></link>
  <meta itemprop="name" content="View Pull Request"></meta>
</div>
<meta itemprop="description" content="View this Pull Request on GitHub"></meta>
</div>

<script type="application/json" data-scope="inboxmarkup">{"api_version":"1.0","publisher":{"api_key":"05dde50f1d1a384dd78767c55493e4bb","name":"GitHub"},"entity":{"external_key":"github/mlpack/mlpack","title":"mlpack/mlpack","subtitle":"GitHub repository","main_image_url":"https://assets-cdn.github.com/images/modules/aws/aws-bg.jpg","avatar_image_url":"https://cloud.githubusercontent.com/assets/143418/15842166/7c72db34-2c0b-11e6-9aed-b52498112777.png","action":{"name":"Open in GitHub","url":"https://github.com/mlpack/mlpack"}},"updates":{"snippets":[{"icon":"PERSON","message":"@MarcosPividori in #726: Continuing with my previous comment:\r\n\r\nI think we have 3 independent decisions to make, when building a binary space tree:\r\n + **Which kind of splitting Hyperplane:** *Axis-Orthogonal*, or *General* (not necessarily Axis-Orthogonal).\r\n + **How to choose the hyperplane:** *Random* or *Widest range* (the one with widest range of projection values).\r\n + **Which Splitting value we wil consider:** *MidPoint value*, *Mean value*, *Median Value*.\r\n\r\nEvery combination is valid, for example:\r\n + *Axis-Orthogonal - Random - MidPoint*.\r\n + *General - Widest range - Mean Value*.\r\n + *Axis-Orthogonal - Widest range - Median Value*.\r\n + etc.\r\n\r\nNow, for Spill Trees, I conside all the combinations between Hyperplane type and Splitting Value, as 2 template parameters. We could add a third template parameter that represents the criterion to choose the hyperplane, between Random and Widest range."}],"action":{"name":"View Pull Request","url":"https://github.com/mlpack/mlpack/pull/726#issuecomment-237876657"}}}</script>