基于优化随机森林算法预测食品检验不合格指标

目的 食品不合格指标危害人类饮食健康，本文将数据挖掘技术应用于食品安全检测中。方法 通过收集山东省食药局官方网站下发的2015~2019年食品安全抽样检验产生的不合格数据，并对其进行多项数据预处理操作，采用超参数网格搜索和10折交叉验证方法建立了基于随机森林的食品不合格指标的分类预测模型，另外，通过对传统随机森林模型的参数优化，将其与决策树（DT）、逻辑回归（LR）和梯度提升决策树（GBDT）算法分类预测结果进行了对比。结果 实验表明经过参数优化后的随机森林模型对食品中不合格指标的预测准确率能够达到89.4%，比DT算法提高了11%，比LR算法提高了9%，比GBDT算法提高了8.1%。结论 基于优化的随机森林模型可以完成食品不合格指标分类预测任务，有广阔的应用前景。

Prediction of unqualified index of food inspection based on optimized random forest algorithm

Objective Food unqualified indicators endanger human dietary health. This paper applies data mining technology to food safety testing.Methods Through the collection of unqualified data generated by the food safety sampling inspection from 2015 to 2019 issued by the official website of Shandong Food and Drug Administration, and a number of data preprocessing operations, the hyperparameter grid search and 10-folds cross-validation method are used to establish A classification prediction model based on random forest-based food unqualified indicators. In addition, by optimizing the parameters of the traditional random forest model, it is classified with decision tree (DT), logistic regression (LR) and gradient boosting decision tree (GBDT) algorithms The forecast results were compared.Results Experiments show that the random forest model after parameter optimization can achieve 89.4% prediction accuracy of unqualified indicators in food, which is 11% higher than the DT algorithm, 9% higher than the LR algorithm, and 8.1% higher than the GBDT algorithm.Conclusion The optimized random forest model can complete the classification and prediction task of food unqualified indicators, and has broad application prospects.