Beef Quality Classification with Reduced E-Nose Data Features According to Beef Cut Types.

Area of Science:

• Food Science
• Analytical Chemistry
• Machine Learning

Background:

• Ensuring global food safety is a significant challenge.
• Monitoring quality and spoilage in short-shelf-life products like beef is difficult.
• Electronic nose technology offers a potential solution for food quality control.

Purpose of the Study:

• To classify beef quality using electronic nose data.
• To develop a novel approach for assessing beef quality and spoilage.
• To evaluate the effectiveness of machine learning models in beef quality classification.

Main Methods:

• Collected electronic nose data from 12 different beef cuts.
• Utilized ANOVA to identify three key active features from 12 initial features.
• Applied Artificial Neural Network (ANN), K Nearest Neighbor (KNN), and Logistic Regression models for classification.
• Classified beef quality into four categories: excellent, good, acceptable, and spoiled.

Main Results:

• Achieved 100% classification accuracy using the Artificial Neural Network (ANN) model.
• The ANN model performed optimally when using combined data from all beef cuts.
• Feature selection using ANOVA identified crucial data points for accurate classification.

Conclusions:

• Electronic nose technology, coupled with machine learning, provides a highly accurate method for assessing beef quality.
• This approach can significantly enhance food safety and reduce spoilage.
• The study demonstrates the potential of ANN for reliable classification of perishable food products.