Risk forewarning model for rice grain Cd pollution based on Bayes theory

Area of Science:

• Agricultural Science
• Environmental Science
• Statistics

Background:

• Cadmium (Cd) contamination in rice grains, originating from polluted soils, poses a significant risk in regions of China.
• Effective risk assessment models are crucial for managing heavy metal contamination in agricultural products.

Purpose of the Study:

• To develop a risk forewarning model for cadmium (Cd) pollution in rice grains using the Bayes classification statistical method.
• To identify key parameters influencing model accuracy and applicability.
• To demonstrate the model's self-renewal capability for improved prediction.

Main Methods:

• Application of the Bayes classification statistical method to establish a risk forewarning model.
• Introduction of 'prior probability factor' and 'data variability factor' into the model.
• Sensitivity analysis to determine the impact of sample size and standard deviation.
• Model self-renewal by incorporating posterior data into prior data for enhanced accuracy.

Main Results:

• The Bayes classification model effectively predicts the risk of cadmium (Cd) pollution in rice grains.
• Model accuracy and applicable range are significantly influenced by sample size and standard deviation.
• Self-renewal of the model by adding posterior data improves prediction accuracy.

Conclusions:

• The Bayes approach provides a feasible method for forewarning heavy metal pollution risks in agricultural products.
• This model can predict Cd pollution risk in rice under specific soil, tillage, and varietal conditions.
• The developed model offers a valuable tool for food safety and agricultural risk management.