• 中国核心期刊(遴选)数据库收录期刊
  • 中文科技期刊数据库收录期刊
  • 中国期刊全文数据库收录期刊
  • 中国学术期刊综合评价数据库统计源期刊等

快速检索引用检索图表检索高级检索

中国医药导刊 ›› 2020, Vol. 22 ›› Issue (12): 883-888.

• 管理与实践 • 上一篇    下一篇

一种面向药品靶向抽检的预测模型

张礼智1, 申亮2*, 王兰霞2, 秦红武1, 李开银3, 陈旺虎1   

  1. 1. 西北师范大学计算机科学与工程学院, 甘肃 兰州 730070;
    2. 甘肃省药品检验研究院, 甘肃 兰州 730070;
    3. 甘肃省食品药品监督管理局, 甘肃 兰州 730070
  • 收稿日期:2020-06-19 修回日期:2020-11-25 出版日期:2020-12-28 发布日期:2020-12-28
  • 基金资助:
    甘肃省食品药品科研项目(项目编号:2018GSFDA050;项目名称:甘肃省食品药品抽检风险预测模型的研究)

A Prediction Model for Targeted Drug Sampling

  1. 1. School of Computer Science and Engineering, Northwest Normal University, Gansu Lanzhou 730070,China;
    2. Gansu Institute for Drug Control, Gansu Lanzhou 730070,China;
    3. Gansu Food and Drug Administration, Gansu Lanzhou 730070,China
  • Received:2020-06-19 Revised:2020-11-25 Online:2020-12-28 Published:2020-12-28

摘要: 药品抽检在药品监管中扮演着十分重要的角色,但有限的监管资源难以满足无限的监管市场。药品的靶向抽检方法可使药品抽检更具针对性,让抽检资源的配置更加高效、合理,从而使有限的抽检资源发挥最大的监督作用。因此,本研究提出了一种支持靶向抽检的药品抽检预测模型。使用某省2013—2018年部分药品抽检历史数据,通过模拟真实的抽检场景,实现了模型的训练与预测评价。根据实验结果,模型的最高预测准确率达到96.6%,不合格药品的预测精确度平均值超过了70%,表明模型能够应用到实际抽检工作当中,可预测下一年度的重点抽检药品,支撑抽检资源的合理调配。
 

关键词: font-size:medium, ">药品检验;靶向抽检;预测模型

Abstract: Drug sampling inspection is critical for drug supervision. However, the limited supervision resources often are difficult to meet the huge supervision requirements. Targeted drug sampling is a feasible measure to solve such a problem. Therefore, a prediction model for targeted sampling is proposed in this paper. Data of drug sampling from 2013 to 2018 based on a province was used to simulate the real sampling scenario, and finish the model training, and the accuracy of the model was evaluated. According to the experiments, the highest prediction accuracy of the model is 96.6% and the average precision of the prediction of unqudified medicine is up to 70%. Therefore, the model can be applied to the real world sampling of medicine to be tested, can predict the key types of medicine to be tested in the next year and optimize the usage of supervision resources.

Key words: font-size:medium, ">Drug test; Targeted sampling; Prediction mode

中图分类号: