Supplementary file 1_Food safety risk prediction and regulatory strategies based on machin

Over 180,000 agricultural product samples collected from Shanghai, Guangzhou, and Shenzhen between January 2023 and March 2025. This dataset supports a machine learning-based risk prediction model developed by Daqing Wu, achieving a recall of 75.4% and precision of 71.9%. Predictive features are ranked based on their contributions to risk, and model interpretability is assessed using SHAP values and probit regression.

Use Cases

• Predicting food safety risk levels for agricultural products based on supply chain stage, region, and product category.
• Analyzing the influence of government supervision intensity on agricultural product safety outcomes.
• Evaluating the impact of weather conditions on food safety risks in megacities.
• Ranking key predictive features for risk using machine learning model contributions.
• Developing interpretable regulatory strategies using SHAP value analysis and econometric methods.

Strengths

• Contains over 180,000 samples, providing a substantial base for model training.
• Model performance metrics are reported: recall of 75.4% and precision of 71.9%.
• Data covers three major Chinese cities (Shanghai, Guangzhou, Shenzhen) over a defined period (Jan 2023-Mar 2025).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is stored in a DOCX file (670.4 KB), which may require extraction of tabular data.

Provenance

Source

figshare, authored by Daqing Wu.

Collection Method

Sampling data collected from Shanghai, Guangzhou, and Shenzhen.

Time Range

January 2023 to March 2025

Freshness

Last updated 2026-04-29 05:57:36; freshness should be verified.

Geography

Shanghai, Guangzhou, Shenzhen, China