Dimensionality reduction

Dimensionality reduction (DR)                                  Machine Learning                  

noun phrase

Definition 1: The transformation of data into a lower-dimensional space while preserving as much structure, variance, or relevant information as possible; common techniques include PCA and t-SNE [Machine Learning Glossary].

Example in context: “Consistent with our approach in machine learning models, we applied dimensionality reduction to the extracted features (i.e., LIWC, TAALES, and BERT) using PCA.” [Tack et al. 2024]

Definition 2: A broader dimensionality-reduction strategy based on reducing the number of variables by selecting a subset of informative features rather than projecting the data into a new space [Bakanach 2023/2024].

Example in context: Dimensionality reduction (DR) has been performed based on two main methods, which are feature selection (FS) and feature extraction (FE).” [Velliangiri et al. 2020]

Synonym: DR

Related terms: feature selection, feature extraction, PCA, t-SNE, latent space, feature space, projection, data compression

Добавить комментарий 0

Ваш электронный адрес не будет опубликован. Обязательные поля помечены *