Assessing the freshness of meat by using quantum-behaved particle swarm optimization and support vector machine

Area of Science:

• Food Science
• Computational Intelligence
• Machine Learning

Background:

• Accurate meat freshness identification is crucial for food safety and quality control.
• Traditional methods relying on single physicochemical or sensory properties often yield suboptimal results.
• Developing robust and accurate meat freshness assessment systems remains a significant challenge.

Purpose of the Study:

• To develop an advanced meat freshness identification system.
• To integrate quantum-behaved particle swarm optimization (QPSO) with support vector machine (SVM) for enhanced performance.
• To accurately classify the freshness of various meat and seafood products.

Main Methods:

• Samples of fresh pork, beef, mutton, and shrimp were stored and monitored over several days.
• Conventional meat freshness indices (total volatile basic nitrogen, aerobic plate count, pH, sensory scores) were measured.
• A support vector machine (SVM) model was developed for freshness assessment.
• Quantum-behaved particle swarm optimization (QPSO) was employed to optimize SVM parameters.

Main Results:

• Individual physicochemical and sensory properties were insufficient for ideal freshness assessment.
• The hybrid SVM model optimized by QPSO demonstrated superior performance.
• The proposed QPSO-SVM model achieved 100% classification accuracy for meat freshness identification.

Conclusions:

• The integration of QPSO and SVM offers a highly effective approach for meat freshness identification.
• This hybrid model overcomes the limitations of traditional assessment methods.
• The QPSO-SVM system provides a reliable tool for accurate and automated meat quality evaluation.