BLOCKCHAIN APPLICATIONS IN BUSINESS OPERATIONS AND SUPPLY CHAIN MANAGEMENT BY MACHINE LEARNING

This study explores the integration of blockchain technology and machine learning (ML) models to improve transparency, efficiency, and resilience in supply chain management. Utilizing a mixed-methods approach, we developed a blockchain framework and tested ML models, including LSTM, ARIMA, Isolation Forest, One-Class SVM, Q-Learning, and Deep Q-Networks, to address demand forecasting, anomaly detection, and optimization. Our findings demonstrate that blockchain significantly enhances data integrity, traceability, and real-time monitoring across supply chains, particularly in industries like food and pharmaceuticals. Among ML models, LSTM showed superior performance for dynamic demand forecasting, while Isolation Forest was highly effective for real-time anomaly detection. Deep Q-Networks excelled in complex optimization tasks but required high computational resources, whereas Q-Learning proved efficient for simpler scenarios. This blockchain-ML integration presents a promising framework for advancing supply chain resilience, enabling secure and agile operations across diverse industrial applications. Limitations include blockchain’s scalability challenges and ML’s computational demands, suggesting areas for future research.