基于人工蜂群优化支持向量机的儿童流感疾病的预测模型研究

Research on the Prediction Model of Child Influenza Disease Based on Artificial Swarm Optimization Support Vector Machine

目的利用人工蜂群优化SVM的预测分析模型(ABC-SVM),改善SVM缺陷的同时提高模型的预测准确度,并运用此模型对儿童流感疾病进行分析预测,为医院提供合理的数据支撑,以更好地应对流感季。方法根据河北省儿童医院在2022~2023年采集的151 d的磷酸奥司他韦颗粒用药量作为研究数据,抽取其中的前101 d作为训练样本,剩余的50 d数据作为测试样本,利用MATLAB R2018b软件编程实现数据的学习与预测,对比预测数据与实际数据的相对误差,且采用决定系数(R2)、均方根误差(RMSE)、平均绝对百分比误差(MAPE)和平均绝对误差(MAE)对所建模型进行对比分析,验证ABC-SVM的预测精度。结果ABC-SVM模型误差的各项指标均最小,且与实际数据相对误差值最小,预测效果最佳。结论本文所建立的ABC-SVM预测模型能预测流感季,对流感控制和预防具有重要意义。

OBJECTIVE To use an improved artificial bee colony(ABC) optimization SVM prediction analysis model to improve SVM defects and improve the prediction accuracy of the model,so as to analyze and predict influenza diseases in children,providing reasonable data support for hospitals to better cope with the upcoming influenza season.METHODS Based on the 151days dose data of oseltamivir collected by children's hospital of Hebei province from 2022 to 2023,the first 101days were selected as training samples,and the remaining 50 days were used as test samples.MATLAB R2018b software was used to program and predict the data.The relative errors between the predicted data and the actual data were compared,and the determination coefficient and root mean square error were used mean absolute percentage error(MAPE) and mean absolute error(MAE) were compared and analyzed to verify the prediction accuracy of ABC-SVM.RESULTST he ABC-SVM model had the smallest indicators of error and the smallest relative error value with actual data,resulting in the best prediction effect.CONCLUSION This indicates that the ABC-SVM prediction model established in this article can accurately predict the arrival of the influenza season,which is of great significance for influenza control and prevention.