<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>Notes on Yuta Ono</title><link>https://yuta0no.github.io/ja/notes/</link><description>Recent content in Notes on Yuta Ono</description><image><title>Yuta Ono</title><url>https://yuta0no.github.io/opengraph.png</url><link>https://yuta0no.github.io/opengraph.png</link></image><generator>Hugo -- 0.146.0</generator><language>ja</language><lastBuildDate>Sat, 16 May 2026 19:00:00 +0900</lastBuildDate><atom:link href="https://yuta0no.github.io/ja/notes/index.xml" rel="self" type="application/rss+xml"/><item><title>指数分布族の自然パラメータに対する経験ベイズ推定値において成り立つ等式</title><link>https://yuta0no.github.io/ja/notes/self-consistency-in-empirical-bayes/</link><pubDate>Sat, 16 May 2026 19:00:00 +0900</pubDate><guid>https://yuta0no.github.io/ja/notes/self-consistency-in-empirical-bayes/</guid><description>十分統計量の期待値が事前分布と事後分布で一致する</description></item><item><title>Forward KLとReverse KL</title><link>https://yuta0no.github.io/ja/notes/forward-reverse-kl/</link><pubDate>Sat, 21 Feb 2026 23:00:00 +0900</pubDate><guid>https://yuta0no.github.io/ja/notes/forward-reverse-kl/</guid><description>KLダイバージェンスは非対称な性質を持つ</description></item><item><title>変分ベイズにおける2通りのELBO導出について</title><link>https://yuta0no.github.io/ja/notes/two-ways-of-elbo-derivation/</link><pubDate>Wed, 30 Jul 2025 17:00:00 +0900</pubDate><guid>https://yuta0no.github.io/ja/notes/two-ways-of-elbo-derivation/</guid><description>ELBO導出においてJensen&amp;rsquo;s inequalityを使うかどうか</description></item><item><title>KLダイバージェンスとFisher情報行列の関係</title><link>https://yuta0no.github.io/ja/notes/kl-divergence-and-fisher-information/</link><pubDate>Wed, 21 May 2025 22:41:00 +0900</pubDate><guid>https://yuta0no.github.io/ja/notes/kl-divergence-and-fisher-information/</guid><description>パラメトリックな確率モデルの微小変化に対するKLダイバージェンスはFisher情報行列を計量行列として計算できる</description></item></channel></rss>