Feature selection Machine Learning
noun phrase
Definition: The process of selecting a subset of the most relevant input features from a dataset for use in model building and training. Its main goals are to reduce irrelevance and redundancy, improve model interpretability, and lower computational cost while preserving or improving predictive performance. Feature selection is related to dimensionality reduction, but it is not identical to it: feature selection retains original variables, whereas dimensionality reduction typically transforms them into a new lower-dimensional representation. [IBM].
Example in context: “Conversely, for non-state-transition modeling, we deploy a feature selection methodology utilizing mutual information (MI) and the Pearson correlation coefficient (PCC) to select appropriate entropy features from several entropy methods.” [Huang et al. 2024]
Related terms: variable selection, attribute selection, dimensionality reduction(related, not identical), feature extraction