Hybrid Machine Learning and Big Data Analytics for Predictive Business Intelligence: A Review of Architectures, Ensemble Methods, and Data-Driven Decision-Making.
Year: 2026
Type: article
Abstract: Organizations generate data faster than traditional reporting tools can analyze it, and business intelligence (BI) has had to change accordingly. This review examines how hybrid machine learning and big data analytics frameworks support predictive BI and data-driven decision-making. It synthesizes five primary studies on intelligent analytics frameworks, AI and data science in IT business intelligence, AI-driven BI architectures, the combination of BI, AI, and big data analytics, and predictive models for strategic decisions, and it draws on foundational work on big data, ensemble learning, and distributed processing. The reviewed literature describes a move from centralized data warehousing to cloud-native architectures and from descriptive reporting to predictive and prescriptive analytics. Hybrid frameworks that combine ensemble methods such as Random Forest and XGBoost with deep learning and traditional approaches are reported to improve forecast accuracy and reliability over individual models. Applications span finance, retail, healthcare, and manufacturing. Open problems include the governance and ethics of AI-driven BI, the interpretability of models, and latency in time-critical settings. We outline future work on human-in-the-loop decision frameworks, edge computing for real-time analytics, and interpretable models.
Source: Figshare
Authors: Shafiqur Rahman, Farhana Afroz
Language: English
Subfield: Management Information Systems
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7230894923
