回帰

• 読み: かいき
• ローマ字: kaiki
• 漢字の分解: 回 (kai) — to return, to revolve; 帰 (ki) — to return home, to revert
• 発音: /ka.i.ki/

意味

Regression; return; recurrence. In statistics, the process of finding the relationship between variables; more broadly, a return to a former state.

A multifaceted noun with both a general meaning (returning to an origin or previous state) and a specific statistical meaning (regression analysis — 回帰分析). In data science and machine learning, 回帰 specifically refers to regression models that predict continuous values, contrasted with 分類 (classification). In biology and medicine, it can refer to the recurrence of a disease. The compound appears in 線形回帰 (linear regression) and 論理回帰 (logistic regression).

例文

1. 機械学習では、回帰モデルを使って気温や株価などの連続値を予測する。
2. 統計的回帰分析により、変数間の相関関係を定量的に明らかにすることができる。
3. 癌の回帰を防ぐため、患者は定期的な検診を受けることが推奨されている。

使い方ガイド

場面: statistics, data science, machine learning, medicine, academic writing

トーン: technical, neutral

起源と歴史

Sino-Japanese compound: 回 (kai) — to return, to revolve + 帰 (ki) — to go back, to return home. Used in classical Japanese to mean 'returning' and adopted in Meiji-era scientific translation to render 'regression' in its statistical sense.

文化的背景

時代: Meiji-Modern

世代: Adult

社会的背景: Educated