<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Conformal Prediction on Nalar</title>
    <link>https://nalar.dev/tags/conformal-prediction/</link>
    <description>Recent content in Conformal Prediction on Nalar</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Fri, 11 Sep 2026 00:00:00 +0700</lastBuildDate>
    <atom:link href="https://nalar.dev/tags/conformal-prediction/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Build Prediction Sets with Conformal Prediction</title>
      <link>https://nalar.dev/build-prediction-sets-with-conformal-prediction/</link>
      <pubDate>Fri, 11 Sep 2026 00:00:00 +0700</pubDate>
      <guid>https://nalar.dev/build-prediction-sets-with-conformal-prediction/</guid>
      <description>&lt;h1 id=&#34;build-prediction-sets-with-conformal-prediction&#34;&gt;Build Prediction Sets with Conformal Prediction&lt;/h1&gt;&#xA;&lt;p&gt;A classifier usually returns one label even when several labels are plausible. That is convenient for software interfaces, but it can hide uncertainty exactly where mistakes are expensive. A document router might be unsure between &lt;code&gt;billing&lt;/code&gt; and &lt;code&gt;account&lt;/code&gt;, yet an &lt;code&gt;argmax&lt;/code&gt; still emits one of them.&lt;/p&gt;&#xA;&lt;p&gt;&lt;strong&gt;Conformal prediction&lt;/strong&gt; offers another interface: return a set of labels sized according to the evidence. Easy inputs can produce one label. Ambiguous inputs can produce several. Under specific assumptions, the procedure also gives a finite-sample coverage guarantee.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
