摘要
为解决医疗机构绩效数据复杂多样,难以快速分析评估的问题,文中提出了一种融合门控运算和LSTM的绩效数据预测模型。该模型以LSTM为核心,加入了循环门控单元GRU,从而可以更好地捕捉绩效历史数据中的长期依赖关系并提高模型训练效率。同时,通过引入自注意力机制可以使其从大量的特征指标中选取关联性较大的特征信息,进一步提高模型的准确性。在公开数据集上进行的实验结果表明,所提模型的MRE仅为3.15%,优于传统机器学习模型。证明该模型能够准确地对绩效数据进行预测分析,为决策的制定提供数据支撑。
In order to solve the problem of complex and diverse performance data in medical institutions,which is difficult to quickly analyze and evaluate,this paper proposes a performance data prediction model that integrates gating operation and LSTM.This model takes LSTM as the core and adds a cyclic gating unit GRU,which can better capture long-term dependencies in performance historical data and improve model training efficiency.At the same time,by introducing a self attention mechanism,highly correlated feature information can be selected from a large number of feature indicators,further improving the accuracy of the model.The experiment results conducted on public datasets show that the MRE of the proposed model is only 3.15%,which is superior to traditional machine learning models.This proves that the model can accurately predict and analyze performance data,providing data support for decision-making.
作者
牛娅敏
张洁
潘利民
郑璐
高雅楠
NIU Ya-min;ZHANG Jie;PAN Li-min;ZHENG Lu;GAO Ya-nan(The First Affiliated Hospital of Hebei North University,Zhangjiakou 075000,Hebei Province,China)
出处
《信息技术》
2025年第6期76-80,共5页
Information Technology
基金
张家口市技术创新引导计划项目(2421149I)。
作者简介
牛娅敏(1985-),女,硕士,会计师,研究方向为网络支付、精细化管理、数字经济。
