<!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>Informationskriterium</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/Informationskriterium"> <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-Informationskriterium rootpage-Informationskriterium 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">Informationskriterium</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>In der <a href="Statistik" title="Statistik">Statistik</a> ist ein <b>Informationskriterium</b> ein <a href="Kriterium" title="Kriterium">Kriterium</a> zur Modellauswahl. Man folgt dabei der Idee von <a href="Ockhams_Rasiermesser" title="Ockhams Rasiermesser">Ockhams Rasiermesser</a>, dass ein Modell nicht unnötig komplex sein soll und balanciert die <a href="Anpassungsg%C3%BCte" title="Anpassungsgüte">Anpassungsgüte</a> des geschätzten Modells an die vorliegenden empirischen Daten (<a href="Stichprobe" title="Stichprobe">Stichprobe</a>) und dessen Komplexität, gemessen an der Anzahl der <a href="Regressionsparameter" title="Regressionsparameter">Parameter</a>, aus. Die Anzahl der Parameter wird dabei „strafend“ berücksichtigt, da sonst komplexe Modelle mit vielen Parametern bevorzugt würden. In diesem Sinne ist das <a href="Korrigiertes_Bestimmtheitsma%C3%9F" class="mw-redirect" title="Korrigiertes Bestimmtheitsmaß">korrigierte Bestimmtheitsmaß</a>, das auf <a href="Henri_Theil" title="Henri Theil">Henri Theil</a> (1970) zurückgeht, ein Vorläufer der heute bekannten Informationskriterien.
</p><p>Allen heute verwendeten Informationskriterien ist gleich, dass sie in zwei verschiedenen Formulierungen vorliegen. Entweder ist das Maß für die Anpassungsgüte als die „<a href="Maximum-Likelihood-Methode" title="Maximum-Likelihood-Methode">maximale Plausibilität</a>“ oder als die „minimale <a href="Varianz_(Stochastik)" title="Varianz (Stochastik)">Varianz</a>“ der <a href="St%C3%B6rgr%C3%B6%C3%9Fe_und_Residuum" title="Störgröße und Residuum">Residuen</a> formuliert. Hieraus ergeben sich unterschiedliche Interpretationsmöglichkeiten. Beim Ersteren ist das Modell „am besten“, bei dem das jeweilige Informationskriterium den höchsten Wert hat (die „strafende“ Anzahl der Parameter muss dabei abgezogen werden). Beim Letzteren ist das Modell mit dem niedrigsten Wert des Informationskriteriums am besten (die Anzahl der Parameter muss „strafend“ addiert werden).
</p>
<div class="mw-heading mw-heading2"><h2 id="Akaike-Informationskriterium"><span id="Akaikes_Informationskriterium"></span> Akaike-Informationskriterium</h2></div>
<p>Das historisch älteste Kriterium wurde im Jahr 1973 von <a href="Hirotsugu_Akaike" title="Hirotsugu Akaike">Hirotsugu Akaike</a> (1927–2009) als <i>an information criterion</i> vorgeschlagen und ist heute als <b>Akaike-Informationskriterium</b><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>, <b>Informationskriterium nach Akaike</b>, oder <b>Akaike'sches Informationskriterium</b> (<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">Akaike information criterion</span>, kurz: <b>AIC</b>) bekannt. Das Akaike-Informationskriterium ist eines der am häufigsten verwendeten Kriterien für die Modellauswahl im Rahmen der Likelihood-basierten Inferenz.
</p><p>In der <a href="Grundgesamtheit" title="Grundgesamtheit">Grundgesamtheit</a> liegt eine Verteilung einer Variablen mit unbekannter <a href="Dichtefunktion" title="Dichtefunktion">Dichtefunktion</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 p}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>p</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle p}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/81eac1e205430d1f40810df36a0edffdc367af36.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; margin-left: -0.089ex; width:1.259ex; height:2.009ex;" alt="{\displaystyle p}" loading="lazy"></span> vor. Bei der <a href="Maximum-Likelihood-Methode" title="Maximum-Likelihood-Methode">Maximum-Likelihood-Schätzung</a> (ML-Schätzung) geht man von einer bekannten Verteilung mit einem unbekannten 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 \theta }">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>θ<!-- θ --></mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \theta }</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/6e5ab2664b422d53eb0c7df3b87e1360d75ad9af.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.09ex; height:2.176ex;" alt="{\displaystyle \theta }" loading="lazy"></span> aus; man nimmt also an, dass sich die Dichtefunktion als <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(\theta )}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>q</mi>
<mo stretchy="false">(</mo>
<mi>θ<!-- θ --></mi>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle q(\theta )}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/f9aab48bedce1a5f1ca74db4f8937032e1f43930.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:3.969ex; height:2.843ex;" alt="{\displaystyle q(\theta )}" loading="lazy"></span> schreiben lässt. Die <a href="Kullback-Leibler-Divergenz" title="Kullback-Leibler-Divergenz">Kullback-Leibler-Divergenz</a> wird als Entfernungsmaß zwischen <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 p}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>p</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle p}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/81eac1e205430d1f40810df36a0edffdc367af36.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; margin-left: -0.089ex; width:1.259ex; height:2.009ex;" alt="{\displaystyle p}" 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 q({\hat {\theta }})}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>q</mi>
<mo stretchy="false">(</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>θ<!-- θ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle q({\hat {\theta }})}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/1c07675c65e53b95602f4ccfa278cbd76213ebb7.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:4.235ex; height:3.343ex;" alt="{\displaystyle q({\hat {\theta }})}" loading="lazy"></span> genutzt. Dabei ist <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 {\theta }}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>θ<!-- θ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\hat {\theta }}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/f0eaae56d74c5844e86caeed8ae205ff9e413bba.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.356ex; height:2.843ex;" alt="{\displaystyle {\hat {\theta }}}" loading="lazy"></span> der geschätzte Parameter aus der Maximum-Likelihood-Schätzung. Je besser das ML-Modell ist, desto kleiner ist die Kullback-Leibler-Divergenz <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(P\|Q)}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>D</mi>
<mo stretchy="false">(</mo>
<mi>P</mi>
<mo fence="false" stretchy="false">‖<!-- ‖ --></mo>
<mi>Q</mi>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle D(P\|Q)}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/2ea3dfe859af4aed411bb763bc399a652433edfc.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:8.48ex; height:2.843ex;" alt="{\displaystyle D(P\|Q)}" loading="lazy"></span>.
</p><p>Für den Fall eines regulären und linearen Modells konnte Akaike zeigen, dass die negative <a href="Log-Likelihood-Funktion" class="mw-redirect" title="Log-Likelihood-Funktion">log-Likelihood-Funktion</a> (auch <i>logarithmische Plausibilitätsfunktion</i> genannt) <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 -\ell ({\hat {\theta }})}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mo>−<!-- − --></mo>
<mi>ℓ<!-- ℓ --></mi>
<mo stretchy="false">(</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>θ<!-- θ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle -\ell ({\hat {\theta }})}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/8bec0abc55aa1bccde2e0083e37691f876c721b7.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:5.943ex; height:3.343ex;" alt="{\displaystyle -\ell ({\hat {\theta }})}" loading="lazy"></span> ein verzerrter Schätzer für die Kullback-Leibler-Divergenz <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(P\|Q)}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>D</mi>
<mo stretchy="false">(</mo>
<mi>P</mi>
<mo fence="false" stretchy="false">‖<!-- ‖ --></mo>
<mi>Q</mi>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle D(P\|Q)}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/2ea3dfe859af4aed411bb763bc399a652433edfc.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:8.48ex; height:2.843ex;" alt="{\displaystyle D(P\|Q)}" loading="lazy"></span> ist und dass die Verzerrung asymptotisch (Stichprobenumfang strebt gegen unendlich) gegen die Zahl der zu schätzenden 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 p}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>p</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle p}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/81eac1e205430d1f40810df36a0edffdc367af36.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; margin-left: -0.089ex; width:1.259ex; height:2.009ex;" alt="{\displaystyle p}" loading="lazy"></span> konvergiert. Für ein Maximum-Likelihood-Modell mit einem <i>p</i>-dimensionalen Parametervektor <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 {\boldsymbol {\theta }}}_{ML}=({\hat {\theta }}_{1},{\hat {\theta }}_{2},\dotsc ,{\hat {\theta }}_{p})^{\mathsf {T}}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">θ<!-- θ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
<mi>L</mi>
</mrow>
</msub>
<mo>=</mo>
<mo stretchy="false">(</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>θ<!-- θ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>1</mn>
</mrow>
</msub>
<mo>,</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>θ<!-- θ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msub>
<mo>,</mo>
<mo>…<!-- … --></mo>
<mo>,</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>θ<!-- θ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>p</mi>
</mrow>
</msub>
<msup>
<mo stretchy="false">)</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="sans-serif">T</mi>
</mrow>
</mrow>
</msup>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\hat {\boldsymbol {\theta }}}_{ML}=({\hat {\theta }}_{1},{\hat {\theta }}_{2},\dotsc ,{\hat {\theta }}_{p})^{\mathsf {T}}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/11e8ff5a571bf05e0c60592483f980293cb4e426.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -1.005ex; width:24.242ex; height:3.509ex;" alt="{\displaystyle {\hat {\boldsymbol {\theta }}}_{ML}=({\hat {\theta }}_{1},{\hat {\theta }}_{2},\dotsc ,{\hat {\theta }}_{p})^{\mathsf {T}}}" loading="lazy"></span>, ist das Akaike-Informationskriterium definiert als<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>
<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 AIC=-2\ell ({\hat {\boldsymbol {\theta }}}_{ML})+2p}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>A</mi>
<mi>I</mi>
<mi>C</mi>
<mo>=</mo>
<mo>−<!-- − --></mo>
<mn>2</mn>
<mi>ℓ<!-- ℓ --></mi>
<mo stretchy="false">(</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">θ<!-- θ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
<mi>L</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
<mo>+</mo>
<mn>2</mn>
<mi>p</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle AIC=-2\ell ({\hat {\boldsymbol {\theta }}}_{ML})+2p}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/91ffa8d7fd1e50aa26cc3491f8f8891eb0e623bc.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:23.237ex; height:3.343ex;" alt="{\displaystyle AIC=-2\ell ({\hat {\boldsymbol {\theta }}}_{ML})+2p}" 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 \ell }">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>ℓ<!-- ℓ --></mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \ell }</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/f066e981e530bacc07efc6a10fa82deee985929e.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:0.97ex; height:2.176ex;" alt="{\displaystyle \ell }" loading="lazy"></span> die log-Likelihood-Funktion darstellt. Das Kriterium ist negativ orientiert, d. h. bei einer Auswahl von möglichen Kandidaten für Modelle (Modellauswahl) für die Daten ist das bevorzugte Modell dasjenige mit dem minimalen AIC-Wert. Das AIC belohnt die <a href="Anpassungsg%C3%BCte" title="Anpassungsgüte">Anpassungsgüte</a> (beurteilt durch die <a href="Likelihood-Funktion" title="Likelihood-Funktion">Likelihood-Funktion</a>), aber es enthält auch einen Strafterm (auch <i>Pönalisierungsterm</i> oder <i>Penalisierungsterm</i> genannt) <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 2p}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mn>2</mn>
<mi>p</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle 2p}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/ff0d3e38652941fc7a2844611f6c7fa8a23104a7.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:2.332ex; height:2.509ex;" alt="{\displaystyle 2p}" loading="lazy"></span>, der hierbei zu hohe Modellkomplexität bestraft. Er ist eine zunehmende Funktion in Abhängigkeit der Anzahl der geschätzten 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 p}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>p</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle p}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/81eac1e205430d1f40810df36a0edffdc367af36.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; margin-left: -0.089ex; width:1.259ex; height:2.009ex;" alt="{\displaystyle p}" loading="lazy"></span>. Der Strafterm verhindert <a href="%C3%9Cberanpassung" title="Überanpassung">Überanpassung</a>, denn die Erhöhung der Anzahl der Parameter im Modell verbessert fast immer die Anpassungsgüte.
Anstelle des AIC nach obiger Definition wird auch <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 AIC/n}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>A</mi>
<mi>I</mi>
<mi>C</mi>
<mrow class="MJX-TeXAtom-ORD">
<mo>/</mo>
</mrow>
<mi>n</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle AIC/n}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/ae9329789a8fdbd46d32abdda3a3719b38246859.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:7.238ex; height:2.843ex;" alt="{\displaystyle AIC/n}" loading="lazy"></span> verwendet, 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 n}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>n</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle n}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/a601995d55609f2d9f5e233e36fbe9ea26011b3b.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.395ex; height:1.676ex;" alt="{\displaystyle n}" loading="lazy"></span> die <a href="Stichprobengr%C3%B6%C3%9Fe" class="mw-redirect" title="Stichprobengröße">Stichprobengröße</a> ist.<sup id="cite_ref-3" class="reference"><a href="#cite_note-3"><span class="cite-bracket">[</span>3<span class="cite-bracket">]</span></a></sup>
</p>
<div class="mw-heading mw-heading3"><h3 id="Allgemeine_Definition">Allgemeine Definition</h3></div>
<p>Angenommen, es liegen <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 n}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>n</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle n}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/a601995d55609f2d9f5e233e36fbe9ea26011b3b.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.395ex; height:1.676ex;" alt="{\displaystyle n}" loading="lazy"></span> unabhängige Beobachtungen mit <a href="Erwartungswert" title="Erwartungswert">Erwartungswert</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 \operatorname {E} (y_{i})=\mu _{i}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi mathvariant="normal">E</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<msub>
<mi>y</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
<mo>=</mo>
<msub>
<mi>μ<!-- μ --></mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \operatorname {E} (y_{i})=\mu _{i}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/b99a8d86897a7015c1412808bd059d2f4c158e36.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:10.631ex; height:2.843ex;" alt="{\displaystyle \operatorname {E} (y_{i})=\mu _{i}}" loading="lazy"></span> und <a href="Varianz_(Stochastik)" title="Varianz (Stochastik)">Varianz</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 \operatorname {Var} (y_{i})=\sigma ^{2}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>Var</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<msub>
<mi>y</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>i</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
<mo>=</mo>
<msup>
<mi>σ<!-- σ --></mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \operatorname {Var} (y_{i})=\sigma ^{2}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/486c20ad0a60fd86433d2d43a8944fcb14f948fa.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:13.049ex; height:3.176ex;" alt="{\displaystyle \operatorname {Var} (y_{i})=\sigma ^{2}}" loading="lazy"></span> vor. Die Variablen <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_{0}=1,x_{1},x_{2},\ldots ,x_{k}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>0</mn>
</mrow>
</msub>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<msub>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>1</mn>
</mrow>
</msub>
<mo>,</mo>
<msub>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msub>
<mo>,</mo>
<mo>…<!-- … --></mo>
<mo>,</mo>
<msub>
<mi>x</mi>
<mrow class="MJX-TeXAtom-ORD">
<mi>k</mi>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle x_{0}=1,x_{1},x_{2},\ldots ,x_{k}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/ac3c4bb8945c781e723cad829a74543ece865250.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:21.077ex; height:2.509ex;" alt="{\displaystyle x_{0}=1,x_{1},x_{2},\ldots ,x_{k}}" loading="lazy"></span> sind verfügbar als potentielle Regressoren. Sei das <a href="Spezifikation_(Statistik)" title="Spezifikation (Statistik)">spezifizierte</a> Modell definiert durch die <a href="Teilmenge" title="Teilmenge">Teilmenge</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 M\subset \{0,1,2,\ldots ,k\}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>M</mi>
<mo>⊂<!-- ⊂ --></mo>
<mo fence="false" stretchy="false">{</mo>
<mn>0</mn>
<mo>,</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 M\subset \{0,1,2,\ldots ,k\}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/cb56b952e2feb6f55b2943ca107493055f1f53b7.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:19.81ex; height:2.843ex;" alt="{\displaystyle M\subset \{0,1,2,\ldots ,k\}}" loading="lazy"></span> von miteinbezogenen erklärenden Variablen mit der dazugehörigen <a href="Versuchsplanmatrix" class="mw-redirect" title="Versuchsplanmatrix">Versuchsplanmatrix</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 \mathbf {X} _{M}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">X</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \mathbf {X} _{M}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/461cbc91d3e8b7fb8966533eb4b65ae83eb5a979.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:3.979ex; height:2.509ex;" alt="{\displaystyle \mathbf {X} _{M}}" loading="lazy"></span>. Für den Kleinste-Quadrate-Schätzer erhält man <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 {\boldsymbol {\beta }}}_{M}=(\mathbf {X} _{M}^{\mathsf {T}}\mathbf {X} _{M})^{-1}\mathbf {X} _{M}^{\mathsf {T}}\mathbf {y} }">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">β<!-- β --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<mo>=</mo>
<mo stretchy="false">(</mo>
<msubsup>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">X</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="sans-serif">T</mi>
</mrow>
</mrow>
</msubsup>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">X</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<msup>
<mo stretchy="false">)</mo>
<mrow class="MJX-TeXAtom-ORD">
<mo>−<!-- − --></mo>
<mn>1</mn>
</mrow>
</msup>
<msubsup>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">X</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="sans-serif">T</mi>
</mrow>
</mrow>
</msubsup>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">y</mi>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\hat {\boldsymbol {\beta }}}_{M}=(\mathbf {X} _{M}^{\mathsf {T}}\mathbf {X} _{M})^{-1}\mathbf {X} _{M}^{\mathsf {T}}\mathbf {y} }</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/99312e2a85e7cb401b3135385b4898d942c62951.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:24.117ex; height:3.343ex;" alt="{\displaystyle {\hat {\boldsymbol {\beta }}}_{M}=(\mathbf {X} _{M}^{\mathsf {T}}\mathbf {X} _{M})^{-1}\mathbf {X} _{M}^{\mathsf {T}}\mathbf {y} }" loading="lazy"></span>.<sup id="cite_ref-4" class="reference"><a href="#cite_note-4"><span class="cite-bracket">[</span>4<span class="cite-bracket">]</span></a></sup>
</p><p>Im Allgemeinen ist das Akaike-Informationskriterium definiert durch
</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 AIC=-2\ell ({\hat {\boldsymbol {\beta }}}_{M},{\hat {\sigma }}^{2};\mathbf {y} ,\mathbf {X} _{M})+2(\mid M\mid +1)}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>A</mi>
<mi>I</mi>
<mi>C</mi>
<mo>=</mo>
<mo>−<!-- − --></mo>
<mn>2</mn>
<mi>ℓ<!-- ℓ --></mi>
<mo stretchy="false">(</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">β<!-- β --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<mo>,</mo>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mo>;</mo>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">y</mi>
</mrow>
<mo>,</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">X</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
<mo>+</mo>
<mn>2</mn>
<mo stretchy="false">(</mo>
<mo>∣<!-- ∣ --></mo>
<mi>M</mi>
<mo>∣<!-- ∣ --></mo>
<mo>+</mo>
<mn>1</mn>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle AIC=-2\ell ({\hat {\boldsymbol {\beta }}}_{M},{\hat {\sigma }}^{2};\mathbf {y} ,\mathbf {X} _{M})+2(\mid M\mid +1)}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/c22602f03041fa901bb5efff60dba933ee0f3d25.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:42.389ex; height:3.343ex;" alt="{\displaystyle AIC=-2\ell ({\hat {\boldsymbol {\beta }}}_{M},{\hat {\sigma }}^{2};\mathbf {y} ,\mathbf {X} _{M})+2(\mid M\mid +1)}" 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 \ell ({\hat {\boldsymbol {\beta }}}_{M},{\hat {\sigma }}^{2})}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>ℓ<!-- ℓ --></mi>
<mo stretchy="false">(</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">β<!-- β --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<mo>,</mo>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \ell ({\hat {\boldsymbol {\beta }}}_{M},{\hat {\sigma }}^{2})}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/2dd65950826ec3d53d84ebf922853251a0fe54b9.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:9.727ex; height:3.343ex;" alt="{\displaystyle \ell ({\hat {\boldsymbol {\beta }}}_{M},{\hat {\sigma }}^{2})}" loading="lazy"></span> der Maximalwert der log-Likelihood-Funktion ist, d. h., die log-Likelihood-Funktion wenn die ML-Schätzer <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 {\boldsymbol {\beta }}}_{M}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">β<!-- β --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\hat {\boldsymbol {\beta }}}_{M}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/4a971f998a3c24fe71f8bbc616b49976f1c6ca85.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:3.53ex; height:3.343ex;" alt="{\displaystyle {\hat {\boldsymbol {\beta }}}_{M}}" 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 {\hat {\sigma }}^{2}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\hat {\sigma }}^{2}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/1ad9d89160c9e63c0aa4c158282cb75a894de56f.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:2.384ex; height:2.676ex;" alt="{\displaystyle {\hat {\sigma }}^{2}}" loading="lazy"></span> in die log-Likelihood-Funktion eingesetzt werden. Kleinere AIC-Werte gehen mit einer besseren Modellanpassung einher. Die Anzahl der Parameter ist hier <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 \mid M\mid +1}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mo stretchy="false">∣<!-- ∣ --></mo>
<mi>M</mi>
<mo>∣<!-- ∣ --></mo>
<mo>+</mo>
<mn>1</mn>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \mid M\mid +1}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/ba82321cfd0dcbace0988f1ed385aa54486d20c6.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:7.997ex; height:2.843ex;" alt="{\displaystyle \mid M\mid +1}" loading="lazy"></span>, da die Varianz der Störgrößen ebenfalls als ein Parameter gezählt wird. In einem <a href="Lineares_Modell" title="Lineares Modell">linearen Modell</a> mit <a href="Normalverteilung" title="Normalverteilung">normalverteilten</a> Störgrößen (<a href="Klassisches_lineares_Modell_der_Normalregression" title="Klassisches lineares Modell der Normalregression">Klassisches lineares Modell der Normalregression</a>) erhält man für die negative log-Likelihood-Funktion (für die Herleitung der log-Likelihood-Funktion, siehe <a href="Klassisches_lineares_Modell_der_Normalregression#Maximum-Likelihood-Schätzung" title="Klassisches lineares Modell der Normalregression">Maximum-Likelihood-Schätzung</a>)
</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 {\begin{aligned}-2\ell ({\hat {\boldsymbol {\beta }}}_{M},{\hat {\sigma }}^{2};\mathbf {y} ,\mathbf {X} _{M})&=-2\ln(L({\hat {\boldsymbol {\beta }}}_{M},{\hat {\sigma }}^{2};\mathbf {y} ,\mathbf {X} _{M}))=n\cdot \ln(2\pi )+n\cdot \ln({\hat {\sigma }}^{2})+{\frac {(\mathbf {y} -\mathbf {X} _{M}{\hat {\boldsymbol {\beta }}}_{M})^{\mathsf {T}}(\mathbf {y} -\mathbf {X} _{M}{\hat {\boldsymbol {\beta }}}_{M})}{{\hat {\sigma }}^{2}}}\\&\propto n\cdot \ln({\hat {\sigma }}^{2})+{\frac {(\mathbf {y} -\mathbf {X} _{M}{\hat {\boldsymbol {\beta }}}_{M})^{\mathsf {T}}(\mathbf {y} -\mathbf {X} _{M}{\hat {\boldsymbol {\beta }}}_{M})}{{\hat {\sigma }}^{2}}}\\&=n\cdot \ln({\hat {\sigma }}^{2})+{\frac {n{\hat {\sigma }}^{2}}{{\hat {\sigma }}^{2}}}\\&=n\cdot \ln({\hat {\sigma }}^{2})+n\\&\propto n\cdot \ln({\hat {\sigma }}^{2})\end{aligned}}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mrow class="MJX-TeXAtom-ORD">
<mtable columnalign="right left right left right left right left right left right left" rowspacing="3pt" columnspacing="0em 2em 0em 2em 0em 2em 0em 2em 0em 2em 0em" displaystyle="true">
<mtr>
<mtd>
<mo>−<!-- − --></mo>
<mn>2</mn>
<mi>ℓ<!-- ℓ --></mi>
<mo stretchy="false">(</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">β<!-- β --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<mo>,</mo>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mo>;</mo>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">y</mi>
</mrow>
<mo>,</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">X</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
</mtd>
<mtd>
<mi></mi>
<mo>=</mo>
<mo>−<!-- − --></mo>
<mn>2</mn>
<mi>ln</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<mi>L</mi>
<mo stretchy="false">(</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">β<!-- β --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<mo>,</mo>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mo>;</mo>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">y</mi>
</mrow>
<mo>,</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">X</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
<mo stretchy="false">)</mo>
<mo>=</mo>
<mi>n</mi>
<mo>⋅<!-- ⋅ --></mo>
<mi>ln</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<mn>2</mn>
<mi>π<!-- π --></mi>
<mo stretchy="false">)</mo>
<mo>+</mo>
<mi>n</mi>
<mo>⋅<!-- ⋅ --></mo>
<mi>ln</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mo stretchy="false">)</mo>
<mo>+</mo>
<mrow class="MJX-TeXAtom-ORD">
<mfrac>
<mrow>
<mo stretchy="false">(</mo>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">y</mi>
</mrow>
<mo>−<!-- − --></mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">X</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">β<!-- β --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<msup>
<mo stretchy="false">)</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="sans-serif">T</mi>
</mrow>
</mrow>
</msup>
<mo stretchy="false">(</mo>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">y</mi>
</mrow>
<mo>−<!-- − --></mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">X</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">β<!-- β --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
</mrow>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
</mfrac>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd></mtd>
<mtd>
<mi></mi>
<mo>∝<!-- ∝ --></mo>
<mi>n</mi>
<mo>⋅<!-- ⋅ --></mo>
<mi>ln</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mo stretchy="false">)</mo>
<mo>+</mo>
<mrow class="MJX-TeXAtom-ORD">
<mfrac>
<mrow>
<mo stretchy="false">(</mo>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">y</mi>
</mrow>
<mo>−<!-- − --></mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">X</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">β<!-- β --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<msup>
<mo stretchy="false">)</mo>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="sans-serif">T</mi>
</mrow>
</mrow>
</msup>
<mo stretchy="false">(</mo>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">y</mi>
</mrow>
<mo>−<!-- − --></mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold">X</mi>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">β<!-- β --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
</mrow>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
</mfrac>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd></mtd>
<mtd>
<mi></mi>
<mo>=</mo>
<mi>n</mi>
<mo>⋅<!-- ⋅ --></mo>
<mi>ln</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mo stretchy="false">)</mo>
<mo>+</mo>
<mrow class="MJX-TeXAtom-ORD">
<mfrac>
<mrow>
<mi>n</mi>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
</mrow>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
</mfrac>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd></mtd>
<mtd>
<mi></mi>
<mo>=</mo>
<mi>n</mi>
<mo>⋅<!-- ⋅ --></mo>
<mi>ln</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mo stretchy="false">)</mo>
<mo>+</mo>
<mi>n</mi>
</mtd>
</mtr>
<mtr>
<mtd></mtd>
<mtd>
<mi></mi>
<mo>∝<!-- ∝ --></mo>
<mi>n</mi>
<mo>⋅<!-- ⋅ --></mo>
<mi>ln</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mo stretchy="false">)</mo>
</mtd>
</mtr>
</mtable>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\begin{aligned}-2\ell ({\hat {\boldsymbol {\beta }}}_{M},{\hat {\sigma }}^{2};\mathbf {y} ,\mathbf {X} _{M})&=-2\ln(L({\hat {\boldsymbol {\beta }}}_{M},{\hat {\sigma }}^{2};\mathbf {y} ,\mathbf {X} _{M}))=n\cdot \ln(2\pi )+n\cdot \ln({\hat {\sigma }}^{2})+{\frac {(\mathbf {y} -\mathbf {X} _{M}{\hat {\boldsymbol {\beta }}}_{M})^{\mathsf {T}}(\mathbf {y} -\mathbf {X} _{M}{\hat {\boldsymbol {\beta }}}_{M})}{{\hat {\sigma }}^{2}}}\\&\propto n\cdot \ln({\hat {\sigma }}^{2})+{\frac {(\mathbf {y} -\mathbf {X} _{M}{\hat {\boldsymbol {\beta }}}_{M})^{\mathsf {T}}(\mathbf {y} -\mathbf {X} _{M}{\hat {\boldsymbol {\beta }}}_{M})}{{\hat {\sigma }}^{2}}}\\&=n\cdot \ln({\hat {\sigma }}^{2})+{\frac {n{\hat {\sigma }}^{2}}{{\hat {\sigma }}^{2}}}\\&=n\cdot \ln({\hat {\sigma }}^{2})+n\\&\propto n\cdot \ln({\hat {\sigma }}^{2})\end{aligned}}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/4a0d4718b78b706fdb4ee311cca38d295627bde5.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -12.838ex; width:105.537ex; height:26.843ex;" alt="{\displaystyle {\begin{aligned}-2\ell ({\hat {\boldsymbol {\beta }}}_{M},{\hat {\sigma }}^{2};\mathbf {y} ,\mathbf {X} _{M})&=-2\ln(L({\hat {\boldsymbol {\beta }}}_{M},{\hat {\sigma }}^{2};\mathbf {y} ,\mathbf {X} _{M}))=n\cdot \ln(2\pi )+n\cdot \ln({\hat {\sigma }}^{2})+{\frac {(\mathbf {y} -\mathbf {X} _{M}{\hat {\boldsymbol {\beta }}}_{M})^{\mathsf {T}}(\mathbf {y} -\mathbf {X} _{M}{\hat {\boldsymbol {\beta }}}_{M})}{{\hat {\sigma }}^{2}}}\\&\propto n\cdot \ln({\hat {\sigma }}^{2})+{\frac {(\mathbf {y} -\mathbf {X} _{M}{\hat {\boldsymbol {\beta }}}_{M})^{\mathsf {T}}(\mathbf {y} -\mathbf {X} _{M}{\hat {\boldsymbol {\beta }}}_{M})}{{\hat {\sigma }}^{2}}}\\&=n\cdot \ln({\hat {\sigma }}^{2})+{\frac {n{\hat {\sigma }}^{2}}{{\hat {\sigma }}^{2}}}\\&=n\cdot \ln({\hat {\sigma }}^{2})+n\\&\propto n\cdot \ln({\hat {\sigma }}^{2})\end{aligned}}}" loading="lazy"></span></dd></dl>
<p>und damit
</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 AIC=n\ln({\hat {\sigma }}^{2})+2(\mid M\mid +1)}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>A</mi>
<mi>I</mi>
<mi>C</mi>
<mo>=</mo>
<mi>n</mi>
<mi>ln</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mo stretchy="false">)</mo>
<mo>+</mo>
<mn>2</mn>
<mo stretchy="false">(</mo>
<mo>∣<!-- ∣ --></mo>
<mi>M</mi>
<mo>∣<!-- ∣ --></mo>
<mo>+</mo>
<mn>1</mn>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle AIC=n\ln({\hat {\sigma }}^{2})+2(\mid M\mid +1)}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/8321423dffaa215e431202a2138490f27a0f9720.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:30.148ex; height:3.176ex;" alt="{\displaystyle AIC=n\ln({\hat {\sigma }}^{2})+2(\mid M\mid +1)}" loading="lazy"></span>.</dd></dl>
<p>Hierbei ist <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 n}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>n</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle n}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/a601995d55609f2d9f5e233e36fbe9ea26011b3b.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.395ex; height:1.676ex;" alt="{\displaystyle n}" loading="lazy"></span> der Stichprobenumfang 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 {\hat {\sigma }}^{2}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\hat {\sigma }}^{2}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/1ad9d89160c9e63c0aa4c158282cb75a894de56f.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:2.384ex; height:2.676ex;" alt="{\displaystyle {\hat {\sigma }}^{2}}" loading="lazy"></span> die Varianz der Störgrößen. Die Varianz der Störgrößen <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 {\sigma }}^{2}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\hat {\sigma }}^{2}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/1ad9d89160c9e63c0aa4c158282cb75a894de56f.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:2.384ex; height:2.676ex;" alt="{\displaystyle {\hat {\sigma }}^{2}}" loading="lazy"></span> wird mittels der <a href="Residuenquadratsumme" title="Residuenquadratsumme">Residuenquadratsumme</a> aus dem Regressionsmodell geschätzt (siehe <a href="Erwartungstreue_Sch%C3%A4tzung_der_Varianz_der_St%C3%B6rgr%C3%B6%C3%9Fen" title="Erwartungstreue Schätzung der Varianz der Störgrößen">Erwartungstreue Schätzung der Varianz der Störgrößen</a>). Allerdings ist zu beachten, dass <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 {\sigma }}^{2}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\hat {\sigma }}^{2}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/1ad9d89160c9e63c0aa4c158282cb75a894de56f.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:2.384ex; height:2.676ex;" alt="{\displaystyle {\hat {\sigma }}^{2}}" loading="lazy"></span> die verzerrte (und nicht wie gewöhnlich die <a href="Erwartungstreue" title="Erwartungstreue">erwartungstreue</a>) Variante der Schätzung der Varianz der Störgrößen <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 {\sigma }}^{2}={\tfrac {1}{n}}{\hat {\boldsymbol {\varepsilon }}}^{\mathsf {T}}{\hat {\boldsymbol {\varepsilon }}}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mo>=</mo>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="false" scriptlevel="0">
<mfrac>
<mn>1</mn>
<mi>n</mi>
</mfrac>
</mstyle>
</mrow>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">ε<!-- ε --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="sans-serif">T</mi>
</mrow>
</mrow>
</msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">ε<!-- ε --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\hat {\sigma }}^{2}={\tfrac {1}{n}}{\hat {\boldsymbol {\varepsilon }}}^{\mathsf {T}}{\hat {\boldsymbol {\varepsilon }}}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/547d51d8b84b336309d9ed6ae03b45e77e9c076b.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -1.005ex; width:11.494ex; height:3.509ex;" alt="{\displaystyle {\hat {\sigma }}^{2}={\tfrac {1}{n}}{\hat {\boldsymbol {\varepsilon }}}^{\mathsf {T}}{\hat {\boldsymbol {\varepsilon }}}}" loading="lazy"></span> ist.<sup id="cite_ref-5" class="reference"><a href="#cite_note-5"><span class="cite-bracket">[</span>5<span class="cite-bracket">]</span></a></sup>
</p>
<div class="mw-heading mw-heading2"><h2 id="Bayessches_Informationskriterium">Bayessches Informationskriterium</h2></div>
<p>Der Nachteil des Akaike-Informationskriteriums ist, dass der Strafterm von der Stichprobengröße unabhängig ist. Bei großen Stichproben sind Verbesserungen der log-Likelihood bzw. der Residualvarianz „leichter“ möglich, weshalb das Kriterium bei großen Stichproben tendenziell Modelle mit verhältnismäßig vielen Parametern vorteilhaft erscheinen lässt. Deshalb empfiehlt sich die Verwendung des durch <a href="Gideon_E._Schwarz" title="Gideon E. Schwarz">Gideon E. Schwarz</a> 1978 vorgeschlagenen <b>bayesschen Informationskriteriums</b><sup id="cite_ref-6" class="reference"><a href="#cite_note-6"><span class="cite-bracket">[</span>6<span class="cite-bracket">]</span></a></sup>, auch <b>Bayes-Informationskriterium</b>, <b>Bayes’sches Informationskriterium</b>, <b>bayesianisches Informationskriterium</b>, oder <b>Schwarz-Bayes-Informationskriterium</b> (kurz: <b>SBC</b>) genannt (<a href="Englische_Sprache" title="Englische Sprache">englisch</a> <i>Bayesian Information Criterion</i>, kurz: <b>BIC</b>). Für ein Modell mit einem Parametervektor <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 {\boldsymbol {\theta }}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold-italic">θ<!-- θ --></mi>
</mrow>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\boldsymbol {\theta }}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/33b025a6bf54ec02e65c871dc3e5897c921419cf.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.306ex; height:2.176ex;" alt="{\displaystyle {\boldsymbol {\theta }}}" loading="lazy"></span>, log-Likelihood-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 \ell ({\boldsymbol {\theta }})}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>ℓ<!-- ℓ --></mi>
<mo stretchy="false">(</mo>
<mrow class="MJX-TeXAtom-ORD">
<mi mathvariant="bold-italic">θ<!-- θ --></mi>
</mrow>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \ell ({\boldsymbol {\theta }})}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/7c36eb40e4f715790a68a8448ba93ec48a47c7af.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:4.085ex; height:2.843ex;" alt="{\displaystyle \ell ({\boldsymbol {\theta }})}" loading="lazy"></span> und dem Maximum-Likelihood-Schätzer <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 {\boldsymbol {\theta }}}_{ML}}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">θ<!-- θ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
<mi>L</mi>
</mrow>
</msub>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle {\hat {\boldsymbol {\theta }}}_{ML}}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/7375dd2eb4aedf163578153b26d7febea0d90260.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:4.535ex; height:3.176ex;" alt="{\displaystyle {\hat {\boldsymbol {\theta }}}_{ML}}" loading="lazy"></span> ist das BIC definiert als<sup id="cite_ref-7" class="reference"><a href="#cite_note-7"><span class="cite-bracket">[</span>7<span class="cite-bracket">]</span></a></sup><sup id="cite_ref-8" class="reference"><a href="#cite_note-8"><span class="cite-bracket">[</span>8<span class="cite-bracket">]</span></a></sup>
</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 BIC=-2\ell ({\hat {\boldsymbol {\theta }}}_{ML})+p\ln(n)}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>B</mi>
<mi>I</mi>
<mi>C</mi>
<mo>=</mo>
<mo>−<!-- − --></mo>
<mn>2</mn>
<mi>ℓ<!-- ℓ --></mi>
<mo stretchy="false">(</mo>
<msub>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi mathvariant="bold-italic">θ<!-- θ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mi>M</mi>
<mi>L</mi>
</mrow>
</msub>
<mo stretchy="false">)</mo>
<mo>+</mo>
<mi>p</mi>
<mi>ln</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<mi>n</mi>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle BIC=-2\ell ({\hat {\boldsymbol {\theta }}}_{ML})+p\ln(n)}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/8d212745b096fe9edc623e3ecee37155780a481a.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:27.626ex; height:3.343ex;" alt="{\displaystyle BIC=-2\ell ({\hat {\boldsymbol {\theta }}}_{ML})+p\ln(n)}" loading="lazy"></span>.</dd></dl>
<p>bzw.
</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 BIC=n\ln({\hat {\sigma }}^{2})+p\ln(n)}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>B</mi>
<mi>I</mi>
<mi>C</mi>
<mo>=</mo>
<mi>n</mi>
<mi>ln</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<msup>
<mrow class="MJX-TeXAtom-ORD">
<mrow class="MJX-TeXAtom-ORD">
<mover>
<mi>σ<!-- σ --></mi>
<mo stretchy="false">^<!-- ^ --></mo>
</mover>
</mrow>
</mrow>
<mrow class="MJX-TeXAtom-ORD">
<mn>2</mn>
</mrow>
</msup>
<mo stretchy="false">)</mo>
<mo>+</mo>
<mi>p</mi>
<mi>ln</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<mi>n</mi>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle BIC=n\ln({\hat {\sigma }}^{2})+p\ln(n)}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/75c1d8471c3126ad191c7450e6446caf5b8f1812.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:25.255ex; height:3.176ex;" alt="{\displaystyle BIC=n\ln({\hat {\sigma }}^{2})+p\ln(n)}" loading="lazy"></span></dd></dl>
<p>Bei diesem Kriterium wächst der Faktor des Strafterms logarithmisch mit der Anzahl der Beobachtungen <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 n}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>n</mi>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle n}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/a601995d55609f2d9f5e233e36fbe9ea26011b3b.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.338ex; width:1.395ex; height:1.676ex;" alt="{\displaystyle n}" loading="lazy"></span>. Bereits ab acht Beobachtungen (<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 \ln 8=2{,}07944>2}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>ln</mi>
<mo><!-- --></mo>
<mn>8</mn>
<mo>=</mo>
<mn>2,079</mn>
<mn>44</mn>
<mo>></mo>
<mn>2</mn>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \ln 8=2{,}07944>2}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/dbf42062d36fab9164d12bcf8cd97e3466550ecc.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.671ex; width:18.47ex; height:2.509ex;" alt="{\displaystyle \ln 8=2{,}07944>2}" loading="lazy"></span>) bestraft das BIC zusätzliche Parameter schärfer als das AIC. Formal ist das BIC identisch zum AIC, bloß dass der Faktor 2 durch <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 \ln(n)}">
<semantics>
<mrow class="MJX-TeXAtom-ORD">
<mstyle displaystyle="true" scriptlevel="0">
<mi>ln</mi>
<mo><!-- --></mo>
<mo stretchy="false">(</mo>
<mi>n</mi>
<mo stretchy="false">)</mo>
</mstyle>
</mrow>
<annotation encoding="application/x-tex">{\displaystyle \ln(n)}</annotation>
</semantics>
</math></span><img src="./_assets_/eb734a37dd21ce173a46342d1cc64c92/b40d8af55c5679aa769abbd67a7b98612c2aeaf5.svg" class="mwe-math-fallback-image-inline mw-invert skin-invert" aria-hidden="true" style="vertical-align: -0.838ex; width:5.143ex; height:2.843ex;" alt="{\displaystyle \ln(n)}" loading="lazy"></span> ersetzt wird.
</p><p>Es hat die gleiche Ausrichtung wie AIC, sodass Modelle mit kleinerem BIC bevorzugt werden.<sup id="cite_ref-9" class="reference"><a href="#cite_note-9"><span class="cite-bracket">[</span>9<span class="cite-bracket">]</span></a></sup>
</p><p>Letzteres Modell wird vor allem in der Soziologie häufig verwendet. Kuha (2004) weist auf die unterschiedlichen Ziele der beiden Kenngrößen hin: Während das BIC versucht dasjenige Modell auszuwählen, das <a href="A-posteriori-Wahrscheinlichkeit" title="A-posteriori-Wahrscheinlichkeit">A-posteriori</a> die größte Plausibilität besitzt das <a href="Wahres_Modell" title="Wahres Modell">wahre Modell</a> zu sein, geht das AIC davon aus, dass es kein wahres Modell gibt.
Die Hälfte des negativen BIC wird auch als Schwarz-Kriterium bezeichnet.
</p>
<div class="mw-heading mw-heading2"><h2 id="Weitere_Informationskriterien">Weitere Informationskriterien</h2></div>
<p>Daneben existieren weitere, seltener verwendete Informationskriterien, wie:
</p>
<ul><li>das Hannan-Quinn-Informationskriterium (<a href="Englische_Sprache" title="Englische Sprache">englisch</a> <i>Hannan-Quinn Information Criterion</i> kurz: <i>HQIC</i>), benannt nach Edward James Hannan und Barry G. Quinn (1979)</li>
<li>das <a href="Deviance_Information_Criterion" title="Deviance Information Criterion">Devianz-Informationskriterium</a> (<a href="Englische_Sprache" title="Englische Sprache">englisch</a> <i>Deviance Information Criterion</i> kurz: <i>DIC</i>), nach Spiegelhalter, Best, Carlin und van der Linde (2002)</li>
<li>Erweitertes Informationskriterium (<a href="Englische_Sprache" title="Englische Sprache">englisch</a> <i>Extended Information Criterion</i>, kurz: <i>EIC</i>) nach Ishiguro, Sakamoto, and Kitagawa (1997)</li>
<li>Fokussiertes Informationskriterium (<a href="Englische_Sprache" title="Englische Sprache">englisch</a> <i>Focused Information Criterion</i>, kurz: <i>FIC</i>) nach Wei (1992), Generalized Information Criterion, kurz: <i>GIC</i> nach Nishii (1984)</li>
<li>Netzwerkinformationskriterium<sup id="cite_ref-10" class="reference"><a href="#cite_note-10"><span class="cite-bracket">[</span>10<span class="cite-bracket">]</span></a></sup> (<a href="Englische_Sprache" title="Englische Sprache">englisch</a> <i>Network Information Criterion</i>, kurz: <i>NIC</i>) nach Murata, Yoshizawa und Amari (1991)</li>
<li>Takeuchi-Informationskriterium (<a href="Englische_Sprache" title="Englische Sprache">englisch</a> <i>Takeuchi's Information Criterion</i>, kurz: <i>TIC</i>) nach Takeuchi (1976)</li></ul>
<p>Ein auf Informationskriterien basierender <a href="Statistischer_Test" title="Statistischer Test">statistischer Test</a> ist der <a href="Vuong-Test" title="Vuong-Test">Vuong-Test</a>.
</p>
<div class="mw-heading mw-heading2"><h2 id="Literatur">Literatur</h2></div>
<ul><li>Hirotsugu Akaike: <i>Information theory and an extension of the maximum likelihood principle.</i> In: B. N. Petrov u. a. (Hrsg.): <i>Proceedings of the Second International Symposium on Information Theory</i> Budapest: Akademiai Kiado 1973. S. 267–281.</li>
<li>Kenneth P. Burnham, David R. Anderson: <i>Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach.</i> Springer-Verlag, New York 2002, ISBN 0-387-95364-7.</li>
<li>Kenneth P. Burnham/David R. Anderson (2004): <i>Multimodel Inference: Understanding AIC and BIC in Model Selection.</i> In: <i>Sociological Methods and Research.</i> Band 33, 2004, <a href="https://doi.org/10.1177/0049124104268644" class="extiw external" title="doi:10.1177/0049124104268644">doi:10.1177/0049124104268644</a>, S. 261–304.</li>
<li>Jouni Kuha (2004): <i>AIC and BIC: Comparisons of Assumptions and Performance</i>, in: <i>Sociological Methods and Research.</i> Band 33, 2004, <a href="https://doi.org/10.1177/0049124103262065" class="extiw external" title="doi:10.1177/0049124103262065">doi:10.1177/0049124103262065</a>, S. 188–229.</li>
<li><a href="Gideon_Schwarz" class="mw-redirect" title="Gideon Schwarz">Gideon Schwarz</a>: <i>Estimating the Dimension of a Model.</i> In: <i>Annals of Statistics.</i> 2, Nr. 6, 1978, <a href="https://doi.org/10.1214/aos/1176344136" class="extiw external" title="doi:10.1214/aos/1176344136">doi:10.1214/aos/1176344136</a>, <a href="JSTOR" title="JSTOR">JSTOR</a>:<a rel="nofollow" class="external text" href="http://www.jstor.org/stable/2958889">2958889</a>, S. 461–464.</li>
<li>David L. Weakliem (2004): <i>Introduction to the Special Issue on Model Selection.</i> In: Sociological Methods and Research, Band 33, 2004, <a href="https://doi.org/10.1177/0049124104268642" class="extiw external" title="doi:10.1177/0049124104268642">doi:10.1177/0049124104268642</a>, S. 167–187.</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"><span class="cite"><a rel="nofollow" class="external text" href="http://isi.cbs.nl/glossary/term81.htm"><i>Akaike's information criterion.</i></a> Glossary of statistical terms. In: <i><a href="International_Statistical_Institute" title="International Statistical Institute">International Statistical Institute</a>.</i> 1. Juni 2011,<span class="Abrufdatum"> abgerufen am 4. Juli 2020</span> (englisch).</span><span style="display: none;" class="Z3988" title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&rfr_id=info%3Asid%2Fde.wikipedia.org%3AInformationskriterium&rft.title=Akaike%27s+information+criterion&rft.description=Akaike%27s+information+criterion&rft.identifier=http%3A%2F%2Fisi.cbs.nl%2Fglossary%2Fterm81.htm&rft.date=2011-06-01&rft.language=en"> </span></span>
</li>
<li id="cite_note-2"><span class="mw-cite-backlink"><a href="#cite_ref-2">↑</a></span> <span class="reference-text"><a href="Ludwig_Fahrmeir" title="Ludwig Fahrmeir">Ludwig Fahrmeir</a>, <a href="Thomas_Kneib" title="Thomas Kneib">Thomas Kneib</a>, Stefan Lang, Brian Marx: <i>Regression: models, methods and applications.</i> Springer Science & Business Media, 2013, ISBN 978-3-642-34332-2, S. 664.</span>
</li>
<li id="cite_note-3"><span class="mw-cite-backlink"><a href="#cite_ref-3">↑</a></span> <span class="reference-text">Ludwig Fahrmeir, Thomas Kneib, Stefan Lang, Brian Marx: <i>Regression: models, methods and applications.</i> Springer Science & Business Media, 2013, ISBN 978-3-642-34332-2, S. 664.</span>
</li>
<li id="cite_note-4"><span class="mw-cite-backlink"><a href="#cite_ref-4">↑</a></span> <span class="reference-text">Ludwig Fahrmeir, Thomas Kneib, Stefan Lang, Brian Marx: <i>Regression: models, methods and applications.</i> Springer Science & Business Media, 2013, ISBN 978-3-642-34332-2, S. 144</span>
</li>
<li id="cite_note-5"><span class="mw-cite-backlink"><a href="#cite_ref-5">↑</a></span> <span class="reference-text">Ludwig Fahrmeir, Thomas Kneib, Stefan Lang, Brian Marx: <i>Regression: models, methods and applications.</i> Springer Science & Business Media, 2013, ISBN 978-3-642-34332-2, S. 148</span>
</li>
<li id="cite_note-6"><span class="mw-cite-backlink"><a href="#cite_ref-6">↑</a></span> <span class="reference-text"><span class="cite"><a rel="nofollow" class="external text" href="http://isi.cbs.nl/glossary/term277.htm"><i>Bayes information criterion.</i></a> Glossary of statistical terms. In: <i><a href="International_Statistical_Institute" title="International Statistical Institute">International Statistical Institute</a>.</i> 1. Juni 2011,<span class="Abrufdatum"> abgerufen am 4. Juli 2020</span> (englisch).</span><span style="display: none;" class="Z3988" title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&rfr_id=info%3Asid%2Fde.wikipedia.org%3AInformationskriterium&rft.title=Bayes+information+criterion&rft.description=Bayes+information+criterion&rft.identifier=http%3A%2F%2Fisi.cbs.nl%2Fglossary%2Fterm277.htm&rft.date=2011-06-01&rft.language=en"> </span></span>
</li>
<li id="cite_note-7"><span class="mw-cite-backlink"><a href="#cite_ref-7">↑</a></span> <span class="reference-text">Leonhard Held und Daniel Sabanés Bové: <a rel="nofollow" class="external text" href="https://www.springer.com/de/book/9783642378867"><i>Applied Statistical Inference: Likelihood and Bayes.</i></a> Springer Heidelberg New York Dordrecht London (2014), ISBN 978-3-642-37886-7, S. 230.</span>
</li>
<li id="cite_note-8"><span class="mw-cite-backlink"><a href="#cite_ref-8">↑</a></span> <span class="reference-text">Ludwig Fahrmeir, <a href="Thomas_Kneib" title="Thomas Kneib">Thomas Kneib</a>, Stefan Lang, Brian Marx: <i>Regression: models, methods and applications.</i> Springer Science & Business Media, 2013, ISBN 978-3-642-34332-2, S. 677.</span>
</li>
<li id="cite_note-9"><span class="mw-cite-backlink"><a href="#cite_ref-9">↑</a></span> <span class="reference-text">Leonhard Held und Daniel Sabanés Bové: <a rel="nofollow" class="external text" href="https://www.springer.com/de/book/9783642378867"><i>Applied Statistical Inference: Likelihood and Bayes.</i></a> Springer Heidelberg New York Dordrecht London (2014), ISBN 978-3-642-37886-7, S. 230.</span>
</li>
<li id="cite_note-10"><span class="mw-cite-backlink"><a href="#cite_ref-10">↑</a></span> <span class="reference-text"><a href="Bastian_Popp_(Wirtschaftswissenschaftler)" title="Bastian Popp (Wirtschaftswissenschaftler)">Bastian Popp</a>: <a rel="nofollow" class="external text" href="https://books.google.de/books?id=M8shBAAAQBAJ&pg=PR19&dq=Netzwerk+Informationskriterium&hl=de&sa=X&ved=0ahUKEwiklLaPwP7kAhXE6aQKHc15C3cQ6AEIMjAB#v=onepage&q=Netzwerk%20Informationskriterium&f=false"><i>Markenerfolg durch Brand Communities: Eine Analyse der Wirkung psychologischer Variablen auf ökonomische Erfolgsindikatoren.</i></a></span>
</li>
</ol>
<div class="mw-heading mw-heading2"><h2 id="Weblinks">Weblinks</h2></div>
<ul><li><span class="cite">Bernard Desgraupes: <a rel="nofollow" class="external text" href="https://cran.r-project.org/web/packages/clusterCrit/vignettes/clusterCrit.pdf"><i>Clustering Indices.</i></a> (PDF) <a href="Universit%C3%A4t_Paris-Nanterre" title="Universität Paris-Nanterre">Universität Paris-Nanterre</a>, März 2013<span style="display:none">;</span><span class="Abrufdatum" style="display:none"> abgerufen am 26. Juni 2016</span> (englisch).</span><span style="display: none;" class="Z3988" title="ctx_ver=Z39.88-2004&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Adc&rfr_id=info%3Asid%2Fde.wikipedia.org%3AInformationskriterium&rft.title=Clustering+Indices&rft.description=Clustering+Indices&rft.identifier=https%3A%2F%2Fcran.r-project.org%2Fweb%2Fpackages%2FclusterCrit%2Fvignettes%2FclusterCrit.pdf&rft.creator=Bernard+Desgraupes&rft.publisher=%5B%5BUniversit%C3%A4t+Paris-Nanterre%5D%5D&rft.date=2013-03&rft.language=en"> </span></li></ul></div><!--htdig_noindex--><div><div class="zim-footer">
Dieser Artikel wurde von <a class="external text" title="Zuletzt bearbeitet am 2025-09-12" href="https://de.wikipedia.org/wiki/?title=Informationskriterium&oldid=259677707">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>