Quantification of Validity and Personalized Decision Framework for Machine Learning Model Repair Strategies Based on Causal Inference
Academic Journal of Computing & Information Science · 2026 · 9 (3)
DOI: 10.25236/ajcis.2026.090302 ↗Kurz gefasst
Fachartikel in der Zeitschrift Academic Journal of Computing & Information Science, 2026.
Abstract (Originalsprache)
: Aiming at the problems of frequent model performance degradation, experience-dependent repair strategies, and lack of quantitative evaluation in machine learning applications, this study constructs a causal inference-driven evaluation and personalized decision framework for model repair based on large-scale controlled data. The average treatment effects are quantified through PSM, IPW, and AIPW, while the heterogeneous treatment effects are estimated by combining causal forests. Effective repair patterns are identified by integrating feature and cost-benefit analysis. This research upgrades model repair from empirical decision-making to data-driven scientific decision-making, provides implementable repair guidelines for industry, and offers a new paradigm for the application of causal inference in machine learning operations and maintenance.
Zitieren
APA
(2026). Quantification of Validity and Personalized Decision Framework for Machine Learning Model Repair Strategies Based on Causal Inference. Academic Journal of Computing & Information Science, 9(3). https://doi.org/10.25236/ajcis.2026.090302
Vancouver
Quantification of Validity and Personalized Decision Framework for Machine Learning Model Repair Strategies Based on Causal Inference. Academic Journal of Computing & Information Science. 2026;9(3). doi:10.25236/ajcis.2026.090302
BibTeX
@article{euvimed2026,
title = {Quantification of Validity and Personalized Decision Framework for Machine Learning Model Repair Strategies Based on Causal Inference},
journal = {Academic Journal of Computing & Information Science},
year = {2026},
volume = {9},
number = {3},
doi = {10.25236/ajcis.2026.090302}
}
EUVIMED