门诊流感样病例少年人数预测模型数据
收藏资源简介：
针对流感样病例监测，基于桐乡各医疗机构门诊就诊记录，获取全年门急诊就诊人数、流感样病例确诊人数、年龄组及性别分布，根据年龄分为婴幼儿、少年、青年、中年、老年，五个年龄层次，筛选出少年年龄层次的门诊流感样病例人数进行预测，根据该年龄层次的人数预测，为少年人群流感防控工作及哨点医院监测工作提供新的思路和方法，并合理配置医疗资源。数据采集：从医疗机构获取匿名化样本统计数据，包括每日门急诊就诊人数、流感样病例症状人数、年龄组及性别分布。数据预处理：数据进行预处理，以消除异常值和噪声，平滑数据，减少随机波动。特征工程：生成针对年龄组大于等于5小于15的特征 AGW_young(t)=α×age_5_15(t)，其中α为模型权重，专门针对5-15岁年龄组。生成性别特征 GW_young(t)，仅针对年龄组大于等于5小于15：GW_young(t)=γ1×male_cases_5_15(t) + γ2×female_cases_5_15(t)其中γ1,γ2为模型权重。模型构建：构建一个专门预测年龄组大于等于5小于15的流感样病例人数的模型 F_young(t)= w1×C_young(t−1)+w2×C_young(t−2)+⋯+wn×C_young(t−n)+β1*AGW_young(t)+β2*GW_young(t)，其中 C_young(t) 代表当天年龄组大于等于5小于15的流感样病例人数；C_young(t−n) 为t−n日年龄组大于等于5小于15的流感样病例人数；wn为历史数据权重；β1和β2是衍生特征权重
For influenza-like illness (ILI) surveillance, based on outpatient visit records from various medical institutions in Tongxiang, this study collects annual outpatient and emergency visit volumes, confirmed ILI case counts, and age and gender distributions. The population is stratified into five age groups: infants, adolescents, young adults, middle-aged adults, and elderly adults. The number of outpatient ILI cases among the adolescent age group is selected for prediction, aiming to provide novel insights and methodologies for influenza prevention and control among adolescents, sentinel hospital surveillance, and rational allocation of medical resources. Data Collection: Anonymized aggregated statistical data is obtained from medical institutions, including daily outpatient and emergency visit volumes, counts of patients with ILI symptoms, and age and gender distributions. Data Preprocessing: The raw data is preprocessed to remove outliers and noise, smooth the dataset, and mitigate random fluctuations. Feature Engineering: A dedicated feature AGW_young(t) is generated for the age group 5 ≤ age < 15, defined as AGW_young(t) = α × age_5_15(t), where α is the model weight. A gender-specific feature GW_young(t) is also generated, exclusively for the 5 ≤ age < 15 age group: GW_young(t) = γ1 × male_cases_5_15(t) + γ2 × female_cases_5_15(t), where γ1 and γ2 are model weights. Model Construction: A dedicated model is developed to predict the number of ILI cases in the 5 ≤ age < 15 age group, formulated as: F_young(t) = w1 × C_young(t−1) + w2 × C_young(t−2) + ⋯ + wn × C_young(t−n) + β1*AGW_young(t) + β2*GW_young(t) In this formula, C_young(t) represents the number of ILI cases in the 5 ≤ age < 15 age group on day t; C_young(t−n) denotes the ILI case count in the same age group on day t−n; wn represents the weight assigned to historical data; β1 and β2 are the weights of the derived features.
