An algorithm for belief rule induction with partial ignorance Li, Yangxue Pérez Gálvez, Ignacio Javier Cabrerizo Lorite, Francisco Javier Morente Molinera, Juan Antonio This work was supported by the grant PID2022-139297OB-I00 funded by MICIU/AEI/10.13039/501100011033 and by “ERDF/EU”, by the China Scholarship Council (CSC) (202106070037), and by the project C-ING-165-UGR23 co-funded by the Regional Ministry of University, Research and Innovation and by the European Union under the Andalusia ERDF Programme 2021–2027. Belief rules are an extension of fuzzy rules that consider belief degrees. They are widely applied due to their traceability, interpretability, and flexibility. However, current belief rules only consider residual support and global ignorance while ignoring partial ignorance. In this paper, we proposed a novel belief rule-based classification system called partial ignorance belief rule induction algorithm (PIBRIA). The consequent part of belief rules takes into account not only the residual support and global ignorance but also partial ignorance of evidence. The belief degrees in the consequent part of each belief rule collectively form a genuine basic probability assignment (BPA). Based on fuzzy unordered rule induction algorithm (FURIA), we propose the learning algorithm for belief rules with partial ignorance, which helps the novel proposed PRIBIA achieve a balance between accuracy and complexity. The inference method employs the Dempster’s combination rule and Shafer’s discounting operation, where the discounting operation can resolve evidence conflicts in the combination rule. The effectiveness and superiority of PIBRIA in handling classification tasks have been validated by comparing it to some state-of-the-art rule-based models. The results of statistical tests show that PIBRIA outperforms all other methods in terms of classification accuracy. Whereas its computational complexity and model complexity are moderate. 2026-01-09T09:58:54Z 2026-01-09T09:58:54Z 2025 journal article Published version: Li, Y., Pérez, I. J., Cabrerizo, F. J., & Morente-Molinera, J. A. (2025). An algorithm for belief rule induction with partial ignorance. Expert Systems with Applications, 261, 125517. https://doi.org/10.1016/j.eswa.2024.125517 https://hdl.handle.net/10481/109370 10.1016/j.eswa.2024.125517 eng http://creativecommons.org/licenses/by-nc-nd/4.0/ open access Attribution-NonCommercial-NoDerivatives 4.0 Internacional Elsevier