Deep-Stacking Network Approach by Multisource Data Mining for Hazardous Risk Identification in IoT-Based Intelligent...

Area of Science:

• Food Science and Technology
• Computer Science and Engineering
• Supply Chain Management

Background:

• Increasing global concerns regarding food quality and safety necessitate advanced solutions.
• Traditional food supply chain risk management faces challenges with real-time data integration.
• Existing real-time risk identification and traceability systems (RITSs) require enhanced predictive capabilities.

Purpose of the Study:

• To develop an innovative deep-stacking network for accurate hazardous risk identification in food supply chains.
• To improve the predictive accuracy and efficiency of real-time risk identification and traceability systems (RITSs).
• To support decision-making for food enterprises, regulatory authorities, and consumers, ensuring food safety and sustainability.

Main Methods:

• Utilized a deep-stacking network approach for hazardous risk identification.
• Integrated massive multisource data from the Internet of Things (IoT) across the entire food supply chain.
• Conducted verification experiments and case analysis to evaluate the method's performance.

Main Results:

• Achieved a prediction accuracy of up to 97.62% for hazardous risk identification.
• Model parameters were optimized to an appropriate size of 211.26 megabytes.
• Demonstrated superior performance in risk level identification compared to existing methods.

Conclusions:

• The proposed deep-stacking network method significantly enhances the capabilities of RITSs for food supply chain security.
• The approach provides accurate, real-time risk assessment, aiding proactive decision-making.
• It fosters collaboration among regulators, enterprises, and consumers for improved food safety and sustainability.