高级检索
    苏海霞, 杨丹凌, 文立, 汤梦莹, 宋晓坤, 黎燕宁. 基于ARIMA与NNAR模型的中国肺癌预测模型构建研究[J]. 广西医科大学学报, 2023, 40(1): 147-153. DOI: 10.16190/j.cnki.45-1211/r.2023.01.023
    引用本文: 苏海霞, 杨丹凌, 文立, 汤梦莹, 宋晓坤, 黎燕宁. 基于ARIMA与NNAR模型的中国肺癌预测模型构建研究[J]. 广西医科大学学报, 2023, 40(1): 147-153. DOI: 10.16190/j.cnki.45-1211/r.2023.01.023
    shu
    Su Haixia, Yang Danling, Wen Li, Tang Mengying, Song Xiaokun, Li Yanning. Study on the construction of lung cancer prediction model in China based on ARIMA and NNAR models[J]. Journal of Guangxi Medical University, 2023, 40(1): 147-153. DOI: 10.16190/j.cnki.45-1211/r.2023.01.023
    Citation: Su Haixia, Yang Danling, Wen Li, Tang Mengying, Song Xiaokun, Li Yanning. Study on the construction of lung cancer prediction model in China based on ARIMA and NNAR models[J]. Journal of Guangxi Medical University, 2023, 40(1): 147-153. DOI: 10.16190/j.cnki.45-1211/r.2023.01.023
    shu

    基于ARIMA与NNAR模型的中国肺癌预测模型构建研究

    Study on the construction of lung cancer prediction model in China based on ARIMA and NNAR models

      摘要
    • 摘要: 目的：基于1990—2019年中国肺癌流行特征数据预测其2020—2024年的发展趋势，为我国肺癌防控相关策略提供科学参考依据。方法：收集1990—2019年中国肺癌（性别）发病率、（性别）死亡率等指标，采用年估计百分比变化（EAPC）分析其变化趋势；比较自回归求和移动平均（ARIMA）模型和神经网络自回归（NNAR）两种模型预测精度，并预测2020—2024年中国肺癌流行趋势。结果：1990—2019年中国肺癌（性别）发病率、（性别）死亡率、（性别）伤残调整寿命年（DALY）率均随时间呈上升趋势；发病率从21.72/10 万增长至58.56/10 万（EAPC 3.72%，P＜0.001）；死亡率从21.65/10 万增长至53.23/10 万（EAPC 3.37%，P＜0.001）；DALY率从588.07/10万增长至1 204.25/10万（EAPC 2.67%，P＜0.001）。ARIMA和NNAR的预测值与实际值基本吻合，ARIMA模型MAPE、MAE、RMSE值更小，预测精度更高。采用ARIMA模型预测得到2020—2024年的发病率为57.67/10 万、59.06/10 万、60.44/10 万、61.83/10 万、63.22/10 万；死亡率分别为53.26/10 万、54.51/10 万、55.76/10 万、57.02/10 万、58.27/10万；DALY率分别为1 191.98/10万、1 211.72/10万、1 231.36/10万、1 250.94/10万、1 270.48/10万。结论：2020—2024年中国肺癌发病、死亡情况仍将加重，ARIMA模型预测中国肺癌流行特征具有较好的精度和预测性能，对肺癌防控策略的制定有指导意义。

       

      Abstract: Objective: To provide scientific references for the prevention and control strategies of lung cancer in China by predicting the development trend of lung cancer in China from 2020 to 2024 based on the epidemiological data of lung cancer in China from 1990 to 2019.Methods: The annual estimated percentage change (EAPC) was used to analyze the trend of incidence (gender), mortality (gender) and other indexes of lung cancer in China from 1990 to 2019.The prediction accuracies of ARIMA model and NNAR model were compared to predict the epidemic trend of lung cancer in China from 2020 to 2024.Results: The incidence (gender), mortality (gender) and disability-adjusted life years (DALY) rate of lung cancer in China increased during the time from 1990 to 2019.The incidence increased from 21.72/100, 000 to 58.56/100, 000 (EAPC 3.72%, P＜ 0.001); the mortality increased from 21.65/100, 000 to 53.23/100, 000 (EAPC 3.37%, P＜ 0.001); the DALY rate increased from 588.07/100, 000 to 1, 204.25/100, 000 (EAPC 2.67%, P＜ 0.001).The predicted values of ARIMA and NNAR were basically consistent with the actual values.The MAPE, MAE and RMSE values of ARIMA model were smaller, and the prediction accuracy was higher.The ARIMA model was used to forecast the incidence, the mortality and the DALY rate.The predictive incidence rates were 57.67/100, 000, 59.06/100, 000, 60.44/100, 000, 61.83/100, 000 and 63.22/100, 000 from 2020 to 2024, respectively; the predictive mortality rates were 53.26/100, 000, 54.51/100, 000, 55.76/100, 000, 57.02/100, 000 and 58.27/100, 000, respectively and the predictive DALY rates were 1, 191.98/100, 000, 1, 211.72/100, 000, 1, 231.36/100, 000, 1, 250.94/100, 000 and 1, 270.48/100, 000, respectively.Conclusion: The incidence and death rate of lung cancer in China will continue to increase from 2020 to 2024.The ARIMA model has good accuracy and predictive performance in predicting the epidemic characteristics of lung cancer in China, which is of guiding significance for the formulation of prevention and control strategies of lung cancer.

       

      • HTML全文
      • 参考文献(0)
      • 相关文章
      • 施引文献
    • 资源附件(0)

    /

    返回文章
    返回