An Empirical Evaluation of Constrained Feature Selection

• Published in: SN computer science Vol. 3; no. 6; p. 445
• Main Authors: Bach, Jakob, Zoller, Kolja, Trittenbach, Holger, Schulz, Katrin, Böhm, Klemens
• Format: Journal Article
• Language: English
• Published: Singapore Springer Nature Singapore 17.08.2022; Springer Nature B.V
• Subjects: Algorithms; Case studies; Computer Imaging; Computer Science; Computer Systems Organization and Communication Networks; Constraints; Data Structures and Information Theory; Datasets; Empirical analysis; Feature selection; Hypotheses; Independent study; Information Systems and Communication Service; Knowledge; Machine learning; Materials science; Original Research; Pattern Recognition and Graphics; Prediction models; Software Engineering/Programming and Operating Systems; Subject specialists; Vision; Feature selection; Constraints; Domain knowledge; Theory-guided data science
• ISSN: 2661-8907; 2662-995X; 2661-8907
• DOI: 10.1007/s42979-022-01338-z

While feature selection helps to get smaller and more understandable prediction models, most existing feature-selection techniques do not consider domain knowledge. One way to use domain knowledge is via constraints on sets of selected features. However, the impact of constraints, e.g., on the predictive quality of selected features, is currently unclear. This article is an empirical study that evaluates the impact of propositional and arithmetic constraints on filter feature selection. First, we systematically generate constraints from various types, using datasets from different domains. As expected, constraints tend to decrease the predictive quality of feature sets, but this effect is non-linear. So we observe feature sets both adhering to constraints and with high predictive quality. Second, we study a concrete setting in materials science. This part of our study sheds light on how one can analyze scientific hypotheses with the help of constraints.