<!DOCTYPE html>
<html class="client-nojs vector-feature-language-in-header-enabled vector-feature-language-in-main-page-header-disabled vector-feature-page-tools-pinned-disabled vector-feature-toc-pinned-clientpref-0 vector-toc-not-available vector-feature-main-menu-pinned-disabled vector-feature-limited-width-clientpref-1 vector-feature-limited-width-content-enabled vector-feature-custom-font-size-clientpref-1 vector-feature-appearance-pinned-clientpref-0 skin-theme-clientpref-day vector-sticky-header-enabled" lang="de" dir="ltr"><head>
<meta charset="UTF-8">
<title>CART (Algorithmus)</title>
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<link rel="icon" type="image/png" href="./_res_/favicon.png">
<link rel="canonical" href="https://de.wikipedia.org/wiki/CART_(Algorithmus)"> <link href="./_mw_/ext.cite.styles.css" rel="stylesheet" type="text/css">
<link href="./_mw_/ext.math.styles.css" rel="stylesheet" type="text/css">
<link href="./_mw_/ext.wikimediamessages.styles.css" rel="stylesheet" type="text/css">
<link href="./_mw_/skins.vector.icons.css" rel="stylesheet" type="text/css">
<link href="./_mw_/skins.vector.search.codex.styles.css" rel="stylesheet" type="text/css">
<link href="./_mw_/skins.vector.styles.css" rel="stylesheet" type="text/css">
<meta name="ResourceLoaderDynamicStyles" content="">
<link href="./_mw_/ext.gadget.citeRef.css" rel="stylesheet" type="text/css">
<link href="./_mw_/ext.gadget.defaultPlainlinks.css" rel="stylesheet" type="text/css">
<link href="./_mw_/ext.gadget.dewikiCommonHide.css" rel="stylesheet" type="text/css">
<link href="./_mw_/ext.gadget.dewikiCommonLayout.css" rel="stylesheet" type="text/css">
<link href="./_mw_/ext.gadget.dewikiCommonStyle.css" rel="stylesheet" type="text/css">
<link href="./_mw_/ext.gadget.dewikiDarkmode.css" rel="stylesheet" type="text/css">
<link href="./_mw_/ext.gadget.dewikiResponsive.css" rel="stylesheet" type="text/css">
<link href="./_mw_/ext.gadget.specialSearch.css" rel="stylesheet" type="text/css">
<link rel="stylesheet" type="text/css" href="./_mw_/site.styles.css">
<link rel="stylesheet" type="text/css" href="./_mw_/noscript.css">
<link rel="stylesheet" type="text/css" href="./_res_/footer.css">
<link rel="stylesheet" type="text/css" href="./_res_/vector-2022.css">
</head>
<body class="skin--responsive skin-vector skin-vector-search-vue mediawiki ltr sitedir-ltr mw-hide-empty-elt ns-0 ns-subject page-CART_Algorithmus rootpage-CART_Algorithmus skin-vector-2022 action-view">
<div class="mw-page-container">
<div class="mw-page-container-inner">
<div class="mw-content-container">
<main id="content" class="mw-body">
<header class="mw-body-header vector-page-titlebar">
<h1 id="firstHeading" class="firstHeading mw-first-heading"><span class="mw-page-title-main">CART (Algorithmus)</span></h1>
</header>
<a id="top"></a>
<div id="bodyContent" class="vector-body ve-init-mw-desktopArticleTarget-targetContainer" aria-labelledby="firstHeading" data-mw-ve-target-container="">
<div id="contentSub">
<div id="mw-content-subtitle"></div>
</div>
<div id="mw-content-text" class="mw-body-content mw-content-ltr" lang="de" dir="ltr"><div class="mw-content-ltr mw-parser-output" lang="de" dir="ltr"><p><b>CART</b> (<i><b>C</b>lassification <b>a</b>nd <b>R</b>egression <b>T</b>rees</i>) ist ein <a href="Algorithmus" title="Algorithmus">Algorithmus</a>, der zur <a href="Entscheidung" title="Entscheidung">Entscheidungsfindung</a> dient. Er wird bei <a href="Entscheidungsbaum" title="Entscheidungsbaum">Entscheidungsbäumen</a> eingesetzt.
</p><p>Der CART-Algorithmus wurde erstmals 1984 von <a href="Leo_Breiman" title="Leo Breiman">Leo Breiman</a> et al. publiziert.<sup id="cite_ref-1" class="reference"><a href="#cite_note-1"><span class="cite-bracket">[</span>1<span class="cite-bracket">]</span></a></sup>
</p>
<div class="mw-heading mw-heading2"><h2 id="Allgemeines">Allgemeines</h2></div>
<p>Ein bedeutendes Merkmal des CART-Algorithmus ist, dass nur <a href="Bin%C3%A4rbaum" title="Binärbaum">Binärbäume</a> erzeugt werden können, das heißt, dass an jeder Verzweigung immer genau zwei Äste vorhanden sind. Das zentrale Element dieses Algorithmus ist also das Finden einer optimalen binären Trennung.
</p><p>Beim CART-Algorithmus wird die Attributsauswahl durch die Maximierung des Informationsgehalts gesteuert. CARTs zeichnen sich dadurch aus, dass sie die Daten in Bezug auf die <a href="Klassifikation" title="Klassifikation">Klassifikation</a> optimal trennen. Dies wird mit einem Schwellwert erreicht, der zu jedem Attribut gesucht wird. Der Informationsgehalt eines Attributes wird als hoch erachtet, wenn durch die Auswertung der sich aus der Teilung über die Schwellwerte ergebenden Attributausprägungen mit einer hohen Trefferquote eine Klassifikation vorgenommen werden kann.
Bei den Entscheidungsbäumen, welche durch den CART-Algorithmus berechnet werden, gilt: Je höher der Informationsgehalt eines Attributs in Bezug auf die Zielgröße, desto weiter oben im Baum findet sich dieses Attribut.
</p><p>Die Entscheidungsschwellwerte ergeben sich jeweils durch die Optimierung der <a href="Entropie_(Informationstheorie)" title="Entropie (Informationstheorie)">Spaltenentropie</a>. Die Gesamtentropien der Attribute ergeben sich durch ein gewichtetes Mittel aus den Spaltenentropien.
</p>
<div class="mw-heading mw-heading2"><h2 id="Mathematische_Beschreibung">Mathematische Beschreibung</h2></div>
<p>Sei <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle {\mathcal {D}}=\{(x^{(i)},y_{i})\}_{i=1}^{m}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">D</mi>
</mrow>
</mrow>
<mo>=</mo>
<mo fence="false" stretchy="false">{</mo>
<mo stretchy="false">(</mo>
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo>,</mo>
<msub>
<mi>y</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
<msubsup>
<mo fence="false" stretchy="false">}</mo>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>m</mi>
</mrow>
</msubsup>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\mathcal {D}}=\{(x^{(i)},y_{i})\}_{i=1}^{m}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/ee5480c7faccf36638eb53bc44b047ab067aa7b2.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -1.005ex; width:18.306ex; height:3.509ex;" alt="{\displaystyle {\mathcal {D}}=\{(x^{(i)},y_{i})\}_{i=1}^{m}}" loading="lazy"></span> die Menge der Trainingsdaten mit Eingabevariablen <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle x^{(i)}=(x_{1}^{(i)},\dots ,x_{n}^{(i)})\in {\mathcal {X}}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo>=</mo>
<mo stretchy="false">(</mo>
<msubsup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>1</mn>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msubsup>
<mo>,</mo>
<mo>…<!-- … --></mo>
<mo>,</mo>
<msubsup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>n</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msubsup>
<mo stretchy="false">)</mo>
<mo>∈<!-- ∈ --></mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">X</mi>
</mrow>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle x^{(i)}=(x_{1}^{(i)},\dots ,x_{n}^{(i)})\in {\mathcal {X}}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/52cd2a1ce6e6a3f73b8c5d9db1489a89616ba17a.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -1.005ex; width:25.028ex; height:3.676ex;" alt="{\displaystyle x^{(i)}=(x_{1}^{(i)},\dots ,x_{n}^{(i)})\in {\mathcal {X}}}" loading="lazy"></span> und Ausgabewerten <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle y\in {\mathcal {Y}}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>y</mi>
<mo>∈<!-- ∈ --></mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">Y</mi>
</mrow>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle y\in {\mathcal {Y}}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/2fca6d018a308729c9ef591e40b7744f1c752c50.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:5.656ex; height:2.509ex;" alt="{\displaystyle y\in {\mathcal {Y}}}" loading="lazy"></span>. Ein Entscheidungsbaum <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle T}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>T</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle T}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/ec7200acd984a1d3a3d7dc455e262fbe54f7f6e0.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.636ex; height:2.176ex;" alt="{\displaystyle T}" loading="lazy"></span> ist formal darstellbar mittels einer Funktion <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle f_{T}:{\mathcal {X}}\rightarrow {\mathcal {Y}}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>f</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>T</mi>
</mrow>
</msub>
<mo>:</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">X</mi>
</mrow>
</mrow>
<mo stretchy="false">→<!-- → --></mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">Y</mi>
</mrow>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle f_{T}:{\mathcal {X}}\rightarrow {\mathcal {Y}}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/2bb30686b35a97a2bccc60db1bbc2188ef706199.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:11.615ex; height:2.509ex;" alt="{\displaystyle f_{T}:{\mathcal {X}}\rightarrow {\mathcal {Y}}}" loading="lazy"></span>, die jeder Eingabe eine Vorhersage des Ausgabewertes zuordnet. Der CART Algorithmus zur Erzeugung eines solchen Baumes findet selbständig die Verzweigungen (<a href="Knoten_(Graphentheorie)" title="Knoten (Graphentheorie)">Knoten</a>) und assoziierten Regeln zur Trennung der Daten (enS <i>split rule</i>), mit denen diese Zuordnung möglichst optimal wird.
</p>
<div class="mw-heading mw-heading3"><h3 id="Regression">Regression</h3></div>
<p>Sei zunächst <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle {\mathcal {Y}}\subseteq \mathbb {R} }">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">Y</mi>
</mrow>
</mrow>
<mo>⊆<!-- ⊆ --></mo>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="double-struck">R</mi>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\mathcal {Y}}\subseteq \mathbb {R} }</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/1bd92d883ccde46c398c619b815038938fbf3d3c.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:6.436ex; height:2.509ex;" alt="{\displaystyle {\mathcal {Y}}\subseteq \mathbb {R} }" loading="lazy"></span>, d. h. die Ausgabe ist ein <a href="Quantit%C3%A4t" title="Quantität">quantitatives</a> Merkmal und der zu konstruierende Baum soll ein <a href="Regressionsanalyse" title="Regressionsanalyse">Regressionsproblem</a> lösen. Um einen optimalen Baum zu finden, muss erst einmal ein Trainingskriterium (<a href="Fehlerfunktion" title="Fehlerfunktion">Fehlerfunktion</a>) definiert werden. Typischerweise wird dafür die <a href="Mittlere_quadratische_Abweichung" title="Mittlere quadratische Abweichung">Mittlere quadratische Abweichung</a> genutzt:
</p>
<dl><dd><span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle E(T,{\mathcal {D}})={\frac {1}{m}}\sum _{i=1}^{m}(f_{T}(x^{(i)})-y_{i})^{2}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>E</mi>
<mo stretchy="false">(</mo>
<mi>T</mi>
<mo>,</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">D</mi>
</mrow>
</mrow>
<mo stretchy="false">)</mo>
<mo>=</mo>
<mrow class="MJX-TeXAtom-ORD">
<mfrac>
<mn>1</mn>
<mi>m</mi>
</mfrac>
</mrow>
<munderover>
<mo>∑<!-- ∑ --></mo>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>m</mi>
</mrow>
</munderover>
<mo stretchy="false">(</mo>
<msub>
<mi>f</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>T</mi>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo stretchy="false">)</mo>
<mo>−<!-- − --></mo>
<msub>
<mi>y</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<msup>
<mo stretchy="false">)</mo>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle E(T,{\mathcal {D}})={\frac {1}{m}}\sum _{i=1}^{m}(f_{T}(x^{(i)})-y_{i})^{2}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/439848547c7a922ffe73b8ca33035ff78d3a17b9.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -3.005ex; width:33.153ex; height:6.843ex;" alt="{\displaystyle E(T,{\mathcal {D}})={\frac {1}{m}}\sum _{i=1}^{m}(f_{T}(x^{(i)})-y_{i})^{2}}" loading="lazy"></span>,</dd></dl>
<p>welche dann minimiert wird. Die Fehlerfunktion ist allerdings generell frei wählbar.
</p><p>Jedem <a href="Bl%C3%A4tter_und_innere_Knoten_in_der_Graphentheorie" title="Blätter und innere Knoten in der Graphentheorie">Blatt</a> <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle l}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>l</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle l}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/829091f745070b9eb97a80244129025440a1cfac.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:0.693ex; height:2.176ex;" alt="{\displaystyle l}" loading="lazy"></span> wird eine Menge <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle R_{l}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle R_{l}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/c9f2439c2bbf45ac826eebcd1bca00db560479be.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:2.486ex; height:2.509ex;" alt="{\displaystyle R_{l}}" loading="lazy"></span> zugeordnet, sodass für alle <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle L}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>L</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle L}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/103168b86f781fe6e9a4a87b8ea1cebe0ad4ede8.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.583ex; height:2.176ex;" alt="{\displaystyle L}" loading="lazy"></span> Blätter die zugeordneten <a href="Disjunkt" title="Disjunkt">disjunkten</a> Mengen eine <a href="Partition_(Mengenlehre)" title="Partition (Mengenlehre)">Partition</a> von <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle D}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>D</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle D}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/f34a0c600395e5d4345287e21fb26efd386990e6.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.924ex; height:2.176ex;" alt="{\displaystyle D}" loading="lazy"></span> bilden. Man sucht nun einen Schätzwert, der für alle <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle x^{(i)}\in R_{l}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo>∈<!-- ∈ --></mo>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle x^{(i)}\in R_{l}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/91ff6f8e8f56563a3b0b3a5c74a34a5fa0219c03.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:8.736ex; height:3.176ex;" alt="{\displaystyle x^{(i)}\in R_{l}}" loading="lazy"></span> möglichst nahe an den wahren Werten <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle \{y_{i}\;|\;x^{(i)}\in R_{l}\}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mo fence="false" stretchy="false">{</mo>
<msub>
<mi>y</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mspace width="thickmathspace"></mspace>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<mspace width="thickmathspace"></mspace>
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo>∈<!-- ∈ --></mo>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mo fence="false" stretchy="false">}</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \{y_{i}\;|\;x^{(i)}\in R_{l}\}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/74dc8feaaa274c3b508e8ef95d0b0f316b9f1e80.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:14.937ex; height:3.343ex;" alt="{\displaystyle \{y_{i}\;|\;x^{(i)}\in R_{l}\}}" loading="lazy"></span> liegt. Der Schätzer
</p>
<dl><dd><span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle f_{T}(x^{(i)}\in R_{l})={\textrm {mean}}(y_{i}\;|\;x^{(i)}\in R_{l})=:{\hat {c}}_{l}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>f</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>T</mi>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo>∈<!-- ∈ --></mo>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
<mo>=</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mtext>mean</mtext>
</mrow>
</mrow>
<mo stretchy="false">(</mo>
<msub>
<mi>y</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mspace width="thickmathspace"></mspace>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<mspace width="thickmathspace"></mspace>
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo>∈<!-- ∈ --></mo>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
<mo>=:</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>c</mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle f_{T}(x^{(i)}\in R_{l})={\textrm {mean}}(y_{i}\;|\;x^{(i)}\in R_{l})=:{\hat {c}}_{l}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/7864ad21a39da0622ba594e775a4fa721daf5762.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:41.776ex; height:3.343ex;" alt="{\displaystyle f_{T}(x^{(i)}\in R_{l})={\textrm {mean}}(y_{i}\;|\;x^{(i)}\in R_{l})=:{\hat {c}}_{l}}" loading="lazy"></span></dd></dl>
<p>liefert dafür eine Lösung. Da die Berechnung aller möglichen Bäume nicht auf effiziente Weise umsetzbar ist, eignet sich für diesen Ansatz ein <a href="Greedy-Algorithmus" title="Greedy-Algorithmus">Greedy-Algorithmus</a> am besten. D.h. konkret: Man beginnt mit einem Baum, der nur aus einem Knoten besteht und findet dann sukzessive lokal optimale Verzweigen. An jedem Knoten bestimmt man das Merkmal <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle j}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>j</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle j}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/2f461e54f5c093e92a55547b9764291390f0b5d0.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; margin-left: -0.027ex; width:0.985ex; height:2.509ex;" alt="{\displaystyle j}" loading="lazy"></span>, das alle Einträge des Elternknoten am besten in zwei Regionen unterteilen kann, wobei immer auch eine optimale Teilungsvorschrift gefunden werden muss. Für <a href="Ordinalskala" title="Ordinalskala">ordinale</a> Merkmale erfolgt dies mittels Schranke <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle s}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>s</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle s}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/01d131dfd7673938b947072a13a9744fe997e632.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.09ex; height:1.676ex;" alt="{\displaystyle s}" loading="lazy"></span>, welche die beiden Regionen <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle R_{1}(j,s)=\{x^{(i)}\;|\;x_{j}^{(i)}\leq s\}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>1</mn>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<mi>j</mi>
<mo>,</mo>
<mi>s</mi>
<mo stretchy="false">)</mo>
<mo>=</mo>
<mo fence="false" stretchy="false">{</mo>
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mspace width="thickmathspace"></mspace>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<mspace width="thickmathspace"></mspace>
<msubsup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>j</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msubsup>
<mo>≤<!-- ≤ --></mo>
<mi>s</mi>
<mo fence="false" stretchy="false">}</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle R_{1}(j,s)=\{x^{(i)}\;|\;x_{j}^{(i)}\leq s\}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/16b3c906021d3e383b87047b206e6687ace1b589.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -1.338ex; width:26.077ex; height:4.009ex;" alt="{\displaystyle R_{1}(j,s)=\{x^{(i)}\;|\;x_{j}^{(i)}\leq s\}}" loading="lazy"></span> und <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle R_{2}(j,s)=\{x^{(i)}\;|\;x_{j}^{(i)}>s\}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<mi>j</mi>
<mo>,</mo>
<mi>s</mi>
<mo stretchy="false">)</mo>
<mo>=</mo>
<mo fence="false" stretchy="false">{</mo>
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mspace width="thickmathspace"></mspace>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<mspace width="thickmathspace"></mspace>
<msubsup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>j</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msubsup>
<mo>></mo>
<mi>s</mi>
<mo fence="false" stretchy="false">}</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle R_{2}(j,s)=\{x^{(i)}\;|\;x_{j}^{(i)}>s\}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/94ac79819fad4a63494784f7d233fd9ffff88095.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -1.338ex; width:26.077ex; height:4.009ex;" alt="{\displaystyle R_{2}(j,s)=\{x^{(i)}\;|\;x_{j}^{(i)}>s\}}" loading="lazy"></span> für alle <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle x^{(i)}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle x^{(i)}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/1496c4d490e7559a75be174f3ead8fedafd6c19c.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:3.409ex; height:2.843ex;" alt="{\displaystyle x^{(i)}}" loading="lazy"></span> in der ursprünglichen Partition des Elternknotens so erzeugt, dass die Fehlerfunktion minimiert wird. Sind die Merkmale nicht ordinal, ergeben sich die Verzweigungsvorschriften anhand der Zuordnung zu den verschiedenen Merkmalsausprägungen. Formal lässt sich das schreiben als
</p>
<dl><dd><span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle \min _{j,s}{\big (}\min _{c_{1}}\sum _{x^{(i)}\in R_{1}(j,s)}(y_{i}-c_{1})^{2}+\min _{c_{2}}\sum _{x^{(i)}\in R_{2}(j,s)}(y_{i}-c_{2})^{2}{\big )}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<munder>
<mo movablelimits="true" form="prefix">min</mo>
<mrow class="MJX-TeXAtom-ORD">
<mi>j</mi>
<mo>,</mo>
<mi>s</mi>
</mrow>
</munder>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mo maxsize="1.2em" minsize="1.2em">(</mo>
</mrow>
</mrow>
<munder>
<mo movablelimits="true" form="prefix">min</mo>
<mrow class="MJX-TeXAtom-ORD">
<msub>
<mi>c</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>1</mn>
</mrow>
</msub>
</mrow>
</munder>
<munder>
<mo>∑<!-- ∑ --></mo>
<mrow class="MJX-TeXAtom-ORD">
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo>∈<!-- ∈ --></mo>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>1</mn>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<mi>j</mi>
<mo>,</mo>
<mi>s</mi>
<mo stretchy="false">)</mo>
</mrow>
</munder>
<mo stretchy="false">(</mo>
<msub>
<mi>y</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mo>−<!-- − --></mo>
<msub>
<mi>c</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>1</mn>
</mrow>
</msub>
<msup>
<mo stretchy="false">)</mo>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mo>+</mo>
<munder>
<mo movablelimits="true" form="prefix">min</mo>
<mrow class="MJX-TeXAtom-ORD">
<msub>
<mi>c</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msub>
</mrow>
</munder>
<munder>
<mo>∑<!-- ∑ --></mo>
<mrow class="MJX-TeXAtom-ORD">
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo>∈<!-- ∈ --></mo>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<mi>j</mi>
<mo>,</mo>
<mi>s</mi>
<mo stretchy="false">)</mo>
</mrow>
</munder>
<mo stretchy="false">(</mo>
<msub>
<mi>y</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mo>−<!-- − --></mo>
<msub>
<mi>c</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msub>
<msup>
<mo stretchy="false">)</mo>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mo maxsize="1.2em" minsize="1.2em">)</mo>
</mrow>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \min _{j,s}{\big (}\min _{c_{1}}\sum _{x^{(i)}\in R_{1}(j,s)}(y_{i}-c_{1})^{2}+\min _{c_{2}}\sum _{x^{(i)}\in R_{2}(j,s)}(y_{i}-c_{2})^{2}{\big )}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/0001d1d378869be51c60d7894fad9723bd2799d2.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -4.005ex; width:55.481ex; height:6.509ex;" alt="{\displaystyle \min _{j,s}{\big (}\min _{c_{1}}\sum _{x^{(i)}\in R_{1}(j,s)}(y_{i}-c_{1})^{2}+\min _{c_{2}}\sum _{x^{(i)}\in R_{2}(j,s)}(y_{i}-c_{2})^{2}{\big )}}" loading="lazy"></span>,</dd></dl>
<p>wobei <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle {\hat {c}}_{k}={\textrm {mean}}(y_{i}\;|\;x^{(i)}\in R_{k}(j,s))_{k=1,2}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>c</mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>k</mi>
</mrow>
</msub>
<mo>=</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mtext>mean</mtext>
</mrow>
</mrow>
<mo stretchy="false">(</mo>
<msub>
<mi>y</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mspace width="thickmathspace"></mspace>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<mspace width="thickmathspace"></mspace>
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo>∈<!-- ∈ --></mo>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>k</mi>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<mi>j</mi>
<mo>,</mo>
<mi>s</mi>
<mo stretchy="false">)</mo>
<msub>
<mo stretchy="false">)</mo>
<mrow class="MJX-TeXAtom-ORD">
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\hat {c}}_{k}={\textrm {mean}}(y_{i}\;|\;x^{(i)}\in R_{k}(j,s))_{k=1,2}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/4024e443ccce3bee8f23579587ef79a68145b81b.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -1.005ex; width:35.05ex; height:3.509ex;" alt="{\displaystyle {\hat {c}}_{k}={\textrm {mean}}(y_{i}\;|\;x^{(i)}\in R_{k}(j,s))_{k=1,2}}" loading="lazy"></span> jeweils die beiden Summen minimiert.
</p><p>Ausgehend von dem einzelnen Knoten, fügt man also in jedem Schritt zwei neue Knoten hinzu, die wiederum solange weiter verzweigt werden, bis eine Abbruchbedingung (z. B. die maximale Pfadlänge von der Wurzel bis zu den Blättern) erfüllt ist.
</p>
<div class="mw-heading mw-heading4"><h4 id="Pruning">Pruning</h4></div>
<p>Da der Baum so in den meisten Fällen zu komplex, also anfällig für <a href="%C3%9Cberanpassung" title="Überanpassung">Überanpassung</a> (<span style="font-style:normal;font-weight:normal"><a href="Englische_Sprache" title="Englische Sprache">englisch</a></span> <span lang="en-Latn" style="font-style:italic">overfitting</span>) ist, kann (sollte) er gestutzt werden (<span style="font-style:normal;font-weight:normal"><a href="Englische_Sprache" title="Englische Sprache">englisch</a></span> <span lang="en-Latn" style="font-style:italic"><a href="Pruning" title="Pruning">Pruning</a></span>). Überanpassung lässt sich verhindern, indem in der Fehlerfunktion ein Regulationsterm (vgl. <span style="font-style:normal;font-weight:normal"><a href="Englische_Sprache" title="Englische Sprache">englisch</a></span> <span lang="en-Latn" style="font-style:italic">Regularization</span>) eingeführt wird, der nicht von den Daten abhängt und die Komplexität des Entscheidungsbaumes bestraft. Dadurch wird unterbunden, dass der Baum spezifische Eigenschaften der Trainingsdaten lernt, die im Allgemeinen (also bei anderen Daten aus der gleichen <a href="Grundgesamtheit" title="Grundgesamtheit">Grundgesamtheit</a>) nicht zu den wahren Vorhersagen beitragen<sup id="cite_ref-2" class="reference"><a href="#cite_note-2"><span class="cite-bracket">[</span>2<span class="cite-bracket">]</span></a></sup>.
</p><p>Die zweite Möglichkeit, die im Folgenden beschrieben wird, ist es zunächst einen relativ großen Baum <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle T_{0}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>T</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>0</mn>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle T_{0}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/55b9e7d7b96196b5a6a26f4349caa3ac82fd67e3.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:2.412ex; height:2.509ex;" alt="{\displaystyle T_{0}}" loading="lazy"></span> zu konstruieren, der dann im Nachhinein beschnitten wird. Sei <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle T\subset T_{0}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>T</mi>
<mo>⊂<!-- ⊂ --></mo>
<msub>
<mi>T</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>0</mn>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle T\subset T_{0}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/53fe5e5032c6af110c4e0d090c36fda33fe45c4d.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:7.147ex; height:2.509ex;" alt="{\displaystyle T\subset T_{0}}" loading="lazy"></span> ein echter <a href="Teilgraph" title="Teilgraph">Teilgraph</a>, der durch Entfernung inneren Knoten erzeugt werden kann (d. h. die Partitionen der Kinder dieses Knotens werden zusammengelegt). Sei <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle |T|}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<mi>T</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle |T|}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/073a0b1ce9d4c91f30329289adbb91aae47f87ea.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:2.93ex; height:2.843ex;" alt="{\displaystyle |T|}" loading="lazy"></span> die Anzahl der Blätter eines solchen Teilgraphs, wobei jedem Blatt <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle l}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>l</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle l}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/829091f745070b9eb97a80244129025440a1cfac.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:0.693ex; height:2.176ex;" alt="{\displaystyle l}" loading="lazy"></span> die Partition <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle R_{l}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle R_{l}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/c9f2439c2bbf45ac826eebcd1bca00db560479be.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:2.486ex; height:2.509ex;" alt="{\displaystyle R_{l}}" loading="lazy"></span> mit <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle |R_{l}|}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle |R_{l}|}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/8fbed57340962a969a093d744fbb26b9519a5506.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:3.78ex; height:2.843ex;" alt="{\displaystyle |R_{l}|}" loading="lazy"></span> Elementen zugeordnet ist. Sei <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle {\hat {c}}_{l}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>c</mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\hat {c}}_{l}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/b6d503d56294fc326e54cb4c11fe8b094934ef23.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:2.014ex; height:2.509ex;" alt="{\displaystyle {\hat {c}}_{l}}" loading="lazy"></span> wie oben und
</p>
<dl><dd><span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle Q_{l}(T,{\mathcal {D_{t}}})={\frac {1}{|R_{l}|}}\sum _{x^{(i)}\in R_{l}}(y_{i}-{\hat {c}}_{l})^{2}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>Q</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<mi>T</mi>
<mo>,</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<msub>
<mi class="MJX-tex-caligraphic" mathvariant="script">D</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">t</mi>
</mrow>
</msub>
</mrow>
</mrow>
<mo stretchy="false">)</mo>
<mo>=</mo>
<mrow class="MJX-TeXAtom-ORD">
<mfrac>
<mn>1</mn>
<mrow>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
</mrow>
</mfrac>
</mrow>
<munder>
<mo>∑<!-- ∑ --></mo>
<mrow class="MJX-TeXAtom-ORD">
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo>∈<!-- ∈ --></mo>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
</mrow>
</munder>
<mo stretchy="false">(</mo>
<msub>
<mi>y</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mo>−<!-- − --></mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>c</mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<msup>
<mo stretchy="false">)</mo>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle Q_{l}(T,{\mathcal {D_{t}}})={\frac {1}{|R_{l}|}}\sum _{x^{(i)}\in R_{l}}(y_{i}-{\hat {c}}_{l})^{2}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/b51a04e31a53729fa75499c21b09538b8a5ddd77.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -3.838ex; width:32.927ex; height:7.176ex;" alt="{\displaystyle Q_{l}(T,{\mathcal {D_{t}}})={\frac {1}{|R_{l}|}}\sum _{x^{(i)}\in R_{l}}(y_{i}-{\hat {c}}_{l})^{2}}" loading="lazy"></span>.</dd></dl>
<p>Die Idee ist es einen Baum zu finden, der die Funktion
</p>
<dl><dd><span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle C_{\alpha }(T,{\mathcal {D_{t}}})=\sum _{l=1}^{|T|}|R_{l}|Q_{l}(T,{\mathcal {D_{t}}})+\alpha |T|}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>C</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>α<!-- α --></mi>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<mi>T</mi>
<mo>,</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<msub>
<mi class="MJX-tex-caligraphic" mathvariant="script">D</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">t</mi>
</mrow>
</msub>
</mrow>
</mrow>
<mo stretchy="false">)</mo>
<mo>=</mo>
<munderover>
<mo>∑<!-- ∑ --></mo>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<mi>T</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
</mrow>
</munderover>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<msub>
<mi>Q</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<mi>T</mi>
<mo>,</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<msub>
<mi class="MJX-tex-caligraphic" mathvariant="script">D</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">t</mi>
</mrow>
</msub>
</mrow>
</mrow>
<mo stretchy="false">)</mo>
<mo>+</mo>
<mi>α<!-- α --></mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<mi>T</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle C_{\alpha }(T,{\mathcal {D_{t}}})=\sum _{l=1}^{|T|}|R_{l}|Q_{l}(T,{\mathcal {D_{t}}})+\alpha |T|}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/13c98a70ef8e433cb85eb1dfbd06b1d3fdcec71c.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -3.005ex; width:37.58ex; height:7.676ex;" alt="{\displaystyle C_{\alpha }(T,{\mathcal {D_{t}}})=\sum _{l=1}^{|T|}|R_{l}|Q_{l}(T,{\mathcal {D_{t}}})+\alpha |T|}" loading="lazy"></span></dd></dl>
<p>für gegebenes <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle \alpha }">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>α<!-- α --></mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \alpha }</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/b79333175c8b3f0840bfb4ec41b8072c83ea88d3.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.488ex; height:1.676ex;" alt="{\displaystyle \alpha }" loading="lazy"></span> minimiert. Hierzu wird eine von <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle {\mathcal {D}}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">D</mi>
</mrow>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\mathcal {D}}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/3277962e1959c3241fb1b70c7f0ac6dcefebd966.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.792ex; height:2.176ex;" alt="{\displaystyle {\mathcal {D}}}" loading="lazy"></span> verschiedene Datenmenge <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle {\mathcal {D_{t}}}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<msub>
<mi class="MJX-tex-caligraphic" mathvariant="script">D</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">t</mi>
</mrow>
</msub>
</mrow>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\mathcal {D_{t}}}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/3cfac976b44bd6d2591dfed770d5f80847c1936c.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:2.618ex; height:2.509ex;" alt="{\displaystyle {\mathcal {D_{t}}}}" loading="lazy"></span> (<span style="font-style:normal;font-weight:normal"><a href="Englische_Sprache" title="Englische Sprache">englisch</a></span> <span lang="en-Latn" style="font-style:italic">test set</span>) benutzt, um einer Überanpassung vorzubeugen (vgl. <a href="Kreuzvalidierungsverfahren" title="Kreuzvalidierungsverfahren">Kreuzvalidierungsverfahren</a>).
</p><p>Der frei wählbare Parameter <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle \alpha }">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>α<!-- α --></mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \alpha }</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/b79333175c8b3f0840bfb4ec41b8072c83ea88d3.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.488ex; height:1.676ex;" alt="{\displaystyle \alpha }" loading="lazy"></span> beschreibt die Gewichtung zwischen Güte der Voraussage des Entscheidungsbaums und seiner Komplexität. Für gegebenes <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle \alpha }">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>α<!-- α --></mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \alpha }</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/b79333175c8b3f0840bfb4ec41b8072c83ea88d3.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.488ex; height:1.676ex;" alt="{\displaystyle \alpha }" loading="lazy"></span> wird eine absteigende Sequenz von Teilbäumen erzeugt, indem bei jedem Schritt der innere Knoten entfernt wird, der den geringsten pro-Knoten Anstieg von <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle \sum \nolimits _{l}|R_{l}|Q_{l}(T,{\mathcal {D_{t}}})}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mo movablelimits="false">∑<!-- ∑ --></mo>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<msub>
<mi>Q</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<mi>T</mi>
<mo>,</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<msub>
<mi class="MJX-tex-caligraphic" mathvariant="script">D</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">t</mi>
</mrow>
</msub>
</mrow>
</mrow>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \sum \nolimits _{l}|R_{l}|Q_{l}(T,{\mathcal {D_{t}}})}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/39ac0e621f18095fb23efbe643049d6d1e0a3269.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -1.338ex; width:17.903ex; height:3.843ex;" alt="{\displaystyle \sum \nolimits _{l}|R_{l}|Q_{l}(T,{\mathcal {D_{t}}})}" loading="lazy"></span> erzeugt, bis nur noch ein einziger Knoten übrig ist. Es gibt dann einen eindeutig bestimmbaren kleinsten Teilbaum, der <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle C_{\alpha }(T,{\mathcal {D_{t}}})}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>C</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>α<!-- α --></mi>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<mi>T</mi>
<mo>,</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<msub>
<mi class="MJX-tex-caligraphic" mathvariant="script">D</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">t</mi>
</mrow>
</msub>
</mrow>
</mrow>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle C_{\alpha }(T,{\mathcal {D_{t}}})}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/b66361ef38c37818ab211d8a2c300d12003154a5.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:10.043ex; height:2.843ex;" alt="{\displaystyle C_{\alpha }(T,{\mathcal {D_{t}}})}" loading="lazy"></span> minimiert.
</p>
<div class="mw-heading mw-heading3"><h3 id="Klassifizierung">Klassifizierung</h3></div>
<p>Seien nun die Ausgabewerte kategorisch, d. h. <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle {\mathcal {Y}}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi class="MJX-tex-caligraphic" mathvariant="script">Y</mi>
</mrow>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\mathcal {Y}}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/8935ef8cf454efa0b58363386c33f16a48ec36ca.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:1.659ex; height:2.509ex;" alt="{\displaystyle {\mathcal {Y}}}" loading="lazy"></span> ist eine endliche Menge und <a href="Ohne_Beschr%C3%A4nkung_der_Allgemeinheit" title="Ohne Beschränkung der Allgemeinheit">o. B. d. A.</a> <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle y_{i}\in \{1,2,\dots ,K\}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>y</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mo>∈<!-- ∈ --></mo>
<mo fence="false" stretchy="false">{</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>,</mo>
<mo>…<!-- … --></mo>
<mo>,</mo>
<mi>K</mi>
<mo fence="false" stretchy="false">}</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle y_{i}\in \{1,2,\dots ,K\}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/e5187004f353aa426e3e1e448c5b3788a8e33a7c.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:17.707ex; height:2.843ex;" alt="{\displaystyle y_{i}\in \{1,2,\dots ,K\}}" loading="lazy"></span>. Die einzigen Änderungen des Algorithmus im Vergleich zur Regression betreffen die Fehlerfunktionen für die Konstruktion des Baums und das Pruning.
</p><p>Wir definieren
</p>
<dl><dd><span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle {\hat {p}}_{lk}:={\frac {1}{|R_{l}|}}\sum _{x_{i}\in R_{l}}I(y_{i}=k)}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>p</mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
<mi>k</mi>
</mrow>
</msub>
<mo>:=</mo>
<mrow class="MJX-TeXAtom-ORD">
<mfrac>
<mn>1</mn>
<mrow>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
</mrow>
</mfrac>
</mrow>
<munder>
<mo>∑<!-- ∑ --></mo>
<mrow class="MJX-TeXAtom-ORD">
<msub>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mo>∈<!-- ∈ --></mo>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
</mrow>
</munder>
<mi>I</mi>
<mo stretchy="false">(</mo>
<msub>
<mi>y</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mo>=</mo>
<mi>k</mi>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\hat {p}}_{lk}:={\frac {1}{|R_{l}|}}\sum _{x_{i}\in R_{l}}I(y_{i}=k)}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/53c263f1c848b67600333c0d76a7d1c21ce98781.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -3.505ex; margin-left: -0.089ex; width:25.864ex; height:6.843ex;" alt="{\displaystyle {\hat {p}}_{lk}:={\frac {1}{|R_{l}|}}\sum _{x_{i}\in R_{l}}I(y_{i}=k)}" loading="lazy"></span>,</dd></dl>
<p>wobei <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle |R_{l}|}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle |R_{l}|}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/8fbed57340962a969a093d744fbb26b9519a5506.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:3.78ex; height:2.843ex;" alt="{\displaystyle |R_{l}|}" loading="lazy"></span> gleich der Anzahl der Elemente in <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle R_{l}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle R_{l}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/c9f2439c2bbf45ac826eebcd1bca00db560479be.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:2.486ex; height:2.509ex;" alt="{\displaystyle R_{l}}" loading="lazy"></span> sei und <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle I(\cdot )}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>I</mi>
<mo stretchy="false">(</mo>
<mo>⋅<!-- ⋅ --></mo>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle I(\cdot )}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/0bbfc685a51919256883dacd3ad86eec3dc7a258.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:3.628ex; height:2.843ex;" alt="{\displaystyle I(\cdot )}" loading="lazy"></span> die <a href="Indikatorfunktion" title="Indikatorfunktion">Indikatorfunktion</a>.
</p><p>Damit lässt sich nun in jeder Region <span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle R_{l}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle R_{l}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/c9f2439c2bbf45ac826eebcd1bca00db560479be.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:2.486ex; height:2.509ex;" alt="{\displaystyle R_{l}}" loading="lazy"></span> die Klassifizierung der Einträge nach Mehrheitsentscheid durchführen:
</p>
<dl><dd><span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle f_{T}(x^{(i)}\in R_{l})={\textrm {argmax}}_{k}\,{\hat {p}}_{lk}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>f</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>T</mi>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<msup>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">(</mo>
<mi>i</mi>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo>∈<!-- ∈ --></mo>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
<mo>=</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mtext>argmax</mtext>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>k</mi>
</mrow>
</msub>
<mspace width="thinmathspace"></mspace>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>p</mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
<mi>k</mi>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle f_{T}(x^{(i)}\in R_{l})={\textrm {argmax}}_{k}\,{\hat {p}}_{lk}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/da5459ac1b592c90946d29deb9aa1bc60f4b71e7.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:28.149ex; height:3.343ex;" alt="{\displaystyle f_{T}(x^{(i)}\in R_{l})={\textrm {argmax}}_{k}\,{\hat {p}}_{lk}}" loading="lazy"></span>.</dd></dl>
<p>Mögliche Fehlerfunktionen geben die sogenannte Unreinheit der Partitionen an und können definiert werden als:
</p>
<dl><dd><span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle {\frac {1}{|R_{l}|}}\sum _{x_{i}\in R_{l}}I(y_{i}\neq k)=1-{\hat {p}}_{lk}\quad }">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mrow class="MJX-TeXAtom-ORD">
<mfrac>
<mn>1</mn>
<mrow>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
<mrow class="MJX-TeXAtom-ORD">
<mo stretchy="false">|</mo>
</mrow>
</mrow>
</mfrac>
</mrow>
<munder>
<mo>∑<!-- ∑ --></mo>
<mrow class="MJX-TeXAtom-ORD">
<msub>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mo>∈<!-- ∈ --></mo>
<msub>
<mi>R</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
</mrow>
</msub>
</mrow>
</munder>
<mi>I</mi>
<mo stretchy="false">(</mo>
<msub>
<mi>y</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mo>≠<!-- ≠ --></mo>
<mi>k</mi>
<mo stretchy="false">)</mo>
<mo>=</mo>
<mn>1</mn>
<mo>−<!-- − --></mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>p</mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
<mi>k</mi>
</mrow>
</msub>
<mspace width="1em"></mspace>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\frac {1}{|R_{l}|}}\sum _{x_{i}\in R_{l}}I(y_{i}\neq k)=1-{\hat {p}}_{lk}\quad }</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/675a6ca01854c97b2bd9237fad8affe65346303c.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -3.505ex; width:31.453ex; height:6.843ex;" alt="{\displaystyle {\frac {1}{|R_{l}|}}\sum _{x_{i}\in R_{l}}I(y_{i}\neq k)=1-{\hat {p}}_{lk}\quad }" loading="lazy"></span> (Missklassifizerungsfehler)</dd>
<dd><span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle \sum _{k=1}^{K}{\hat {p}}_{lk}(1-{\hat {p}}_{lk})\quad }">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<munderover>
<mo>∑<!-- ∑ --></mo>
<mrow class="MJX-TeXAtom-ORD">
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>K</mi>
</mrow>
</munderover>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>p</mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
<mi>k</mi>
</mrow>
</msub>
<mo stretchy="false">(</mo>
<mn>1</mn>
<mo>−<!-- − --></mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>p</mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
<mi>k</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
<mspace width="1em"></mspace>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \sum _{k=1}^{K}{\hat {p}}_{lk}(1-{\hat {p}}_{lk})\quad }</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/88045ae63586653f7bc890db2ae289bd669a3cb8.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -3.005ex; width:17.754ex; height:7.343ex;" alt="{\displaystyle \sum _{k=1}^{K}{\hat {p}}_{lk}(1-{\hat {p}}_{lk})\quad }" loading="lazy"></span> (<a href="Simpson-Index" title="Simpson-Index">Gini-Simpson-Index</a>)</dd>
<dd><span class="mwe-math-element mwe-math-element-inline"><span class="mwe-math-mathml-inline mwe-math-mathml-a11y" style="display: none;"><math xmlns="http://www.w3.org/1998/Math/MathML" alttext="{\displaystyle -\sum _{k=1}^{K}{\hat {p}}_{lk}\log {\hat {p}}_{lk}\quad }">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mo>−<!-- − --></mo>
<munderover>
<mo>∑<!-- ∑ --></mo>
<mrow class="MJX-TeXAtom-ORD">
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>K</mi>
</mrow>
</munderover>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>p</mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
<mi>k</mi>
</mrow>
</msub>
<mi>log</mi>
<mo><!-- --></mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>p</mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>l</mi>
<mi>k</mi>
</mrow>
</msub>
<mspace width="1em"></mspace>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle -\sum _{k=1}^{K}{\hat {p}}_{lk}\log {\hat {p}}_{lk}\quad }</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/0557e161eef97b0b1ad24a7ea35f888fe162cef8.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -3.005ex; width:17.883ex; height:7.343ex;" alt="{\displaystyle -\sum _{k=1}^{K}{\hat {p}}_{lk}\log {\hat {p}}_{lk}\quad }" loading="lazy"></span> (<a href="Kreuzentropie" title="Kreuzentropie">Kreuzentropie</a>, <a href="Shannon-Index" title="Shannon-Index">Shannon-Index</a>)</dd></dl>
<p>Jede dieser Funktionen kann bei der Konstruktion eines Entscheidungsbaumes anstelle der mittleren quadratischen Abweichung genutzt werden, sodass der wesentliche Teil des Algorithmus unverändert bleibt.
</p>
<div class="mw-heading mw-heading2"><h2 id="Siehe_auch">Siehe auch</h2></div>
<ul><li><a href="Iterative_Dichotomiser_3" title="Iterative Dichotomiser 3">Iterative Dichotomiser 3</a> (ID3)</li>
<li><a href="C4.5" title="C4.5">C4.5</a></li>
<li><a href="CHAID" title="CHAID">CHAID</a></li>
<li><a href="Klassifikationsbaum-Methode" title="Klassifikationsbaum-Methode">Klassifikationsbaum-Methode</a></li>
<li>Implementierungen
<ul><li>In <a href="R_(Programmiersprache)" title="R (Programmiersprache)">R</a>: <a rel="nofollow" class="external text" href="https://www.rdocumentation.org/packages/rpart/versions/4.1-15">rpart</a></li>
<li>In <a href="Python_(Programmiersprache)" title="Python (Programmiersprache)">Python</a> mit <a href="Scikit-learn" title="Scikit-learn">Scikit-learn</a>: <a rel="nofollow" class="external text" href="https://scikit-learn.org/stable/modules/tree.html#tree-algorithms">sklearn.tree</a></li></ul></li></ul>
<div class="mw-heading mw-heading2"><h2 id="Literatur">Literatur</h2></div>
<ul><li>T. Hastie, R. Tibshirani, J. Friedman: <a rel="nofollow" class="external text" href="https://web.stanford.edu/~hastie/ElemStatLearn/printings/ESLII_print12.pdf"><i>The Elements of Statistical Learning: Data Mining, Inference, and Prediction.</i></a> Second Edition. Springer-Verlag New York, 2009, ISBN 978-0-387-84857-0</li></ul>
<div class="mw-heading mw-heading2"><h2 id="Einzelnachweise">Einzelnachweise</h2></div>
<ol class="references">
<li id="cite_note-1"><span class="mw-cite-backlink"><a href="#cite_ref-1">↑</a></span> <span class="reference-text">L. Breiman, J. H. Friedman, R. A. Olshen, C. J. Stone: <i>CART: Classification and Regression Trees</i>. Wadsworth: Belmont, CA, 1984.</span>
</li>
<li id="cite_note-2"><span class="mw-cite-backlink"><a href="#cite_ref-2">↑</a></span> <span class="reference-text">T. Grubinger, A. Zeileis, K.-P. Pfeiffer: <a rel="nofollow" class="external text" href="https://www.jstatsoft.org/article/view/v061i01"><i>evtree: Evolutionary Learning of Globally Optimal Classification and Regression Trees in R</i></a>. Journal of Statistical Software. 2014, Volume 61, Issue 1</span>
</li>
</ol></div><!--htdig_noindex--><div><div class="zim-footer">
Dieser Artikel wurde von <a class="external text" title="Zuletzt bearbeitet am 2025-05-17" href="https://de.wikipedia.org/wiki/?title=CART_(Algorithmus)&oldid=256091557">Wikipedia</a> herausgegeben. Der Text ist unter <a class="external text" href="https://creativecommons.org/licenses/by-sa/4.0/deed.de">Creative Commons Attribution-Share Alike 4.0</a> verfügbar, sofern nicht anders angegeben. Für die Mediendateien können zusätzliche Bedingungen gelten.
</div>
</div><!--/htdig_noindex--></div>
</div>
</main>
</div>
</div>
</div>
<script src="./_webp_/webpHandler.js"></script>
</body></html>