Abstract
The paper presents a research methodology focused on generating a global attribute ranking, based on discrete variants of datasets, transformed by multiple algorithms. The approach enables to accumulate information on feature importance from such local sources and, when it is represented in the form of a global ranking, identify the features that are the most relevant for decision-making processes. The research procedure was validated by experiments, in which the rankings were used to control filtering decision rules induced by the classic rough set approach. In the vast majority of cases, it was possible to obtain noticeable attribute reduction, and predictions improved or comparable with the results obtained for local variants of the data.
Paper Type
Short Paper
DOI
10.62036/ISD.2024.29
Construction of Features Ranking— Global Approach
The paper presents a research methodology focused on generating a global attribute ranking, based on discrete variants of datasets, transformed by multiple algorithms. The approach enables to accumulate information on feature importance from such local sources and, when it is represented in the form of a global ranking, identify the features that are the most relevant for decision-making processes. The research procedure was validated by experiments, in which the rankings were used to control filtering decision rules induced by the classic rough set approach. In the vast majority of cases, it was possible to obtain noticeable attribute reduction, and predictions improved or comparable with the results obtained for local variants of the data.
Recommended Citation
Zielosko, B., Stańczyk, U. & Jabloński, K. (2024). Construction of Features Ranking— Global Approach. In B. Marcinkowski, A. Przybylek, A. Jarzębowicz, N. Iivari, E. Insfran, M. Lang, H. Linger, & C. Schneider (Eds.), Harnessing Opportunities: Reshaping ISD in the post-COVID-19 and Generative AI Era (ISD2024 Proceedings). Gdańsk, Poland: University of Gdańsk. ISBN: 978-83-972632-0-8. https://doi.org/10.62036/ISD.2024.29