EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Integrating hybrid machine learning and Bayesian inference for predictive risk management in business organizations

Irfan Mohd

Digital Enterprise Studies · 2026

DOI: 10.68217/pr0x6441 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Digital Enterprise Studies, 2026. Verfasst von Irfan Mohd. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Business organizations increasingly use predictive analytics to identify emerging operational, financial, and strategic risks, yet predictions alone do not establish how uncertainty should be represented or translated into risk decisions. This conceptual article develops an integrated framework linking hybrid machine learning, Bayesian inference, and organizational risk management. Drawing on an integrative synthesis of research in enterprise risk management, probabilistic machine learning, Bayesian decision theory, organizational information processing, and dynamic capabilities, it distinguishes three mechanisms, namely predictive signal extraction, evidence-based belief updating, and loss-sensitive decision support. It specifies two alternative architectures, one placing Bayesian estimation within the predictive model and one using a conventional model output as evidence through a conditional evidence model estimated on held-out or cross-fitted predictions, and treats Bayesian dependency graphs as a structure compatible with either rather than a third option. The article separates probability calibration from likelihood estimation, states when overlapping prior and evidence information can be combined without double counting, and explains how data quality, prior elicitation, monitoring, managerial comprehension, and governance condition the usefulness of the resulting risk information. Six conditional propositions link these mechanisms to risk information quality, adaptive assessment, and decision quality. The framework explains when integrated systems may support risk management without assuming that Bayesian inference improves accuracy, calibration, or interpretability. A research agenda identifies matched benchmarks, comprehension-based decision experiments, and longitudinal organizational studies.

Zitieren

APA

Mohd, I. (2026). Integrating hybrid machine learning and Bayesian inference for predictive risk management in business organizations. Digital Enterprise Studies. https://doi.org/10.68217/pr0x6441

Vancouver

Mohd I. Integrating hybrid machine learning and Bayesian inference for predictive risk management in business organizations. Digital Enterprise Studies. 2026. doi:10.68217/pr0x6441

BibTeX
@article{Mohd2026,
  author = {Mohd, Irfan},
  title = {Integrating hybrid machine learning and Bayesian inference for predictive risk management in business organizations},
  journal = {Digital Enterprise Studies},
  year = {2026},
  doi = {10.68217/pr0x6441}
}
Kostenlos starten →