Hanamaru 日本語にほんご × ML/AI
← All topics

Core Machine Learning Topics

Feature Engineering

特徴量とくちょうりょうエンジニアリング

Feature engineering improves model inputs through domain-informed transformations. Feature selection keeps useful inputs, while feature extraction derives informative representations from raw data.

Japanese terms

  1. Feature engineering — 特徴量とくちょうりょうエンジニアリング: The process of creating or transforming model inputs to improve learning and performance.
  2. Feature selection — 特徴選択とくちょうせんたく: Selecting a useful subset of available input features.
  3. Feature extraction — 特徴抽出とくちょうちゅうしゅつ: Transforming raw data into informative derived features.

Data and datasets supplies the raw material, and embeddings and representation learning automate many feature-learning tasks.