太原师范学院学报(自然科学版)

使用“基于决策类划分的新型多变量决策树算法”为COVID-19 Surveillance数据集构建决策逻辑,成功推导出两种等效且分类准确的决策表达式,将病情分为轻度、中度和重度,所得结果与我国诊疗方案类似.新算法在时间复杂度上优于其他两种算法,在处理7个条件属性的数据集时表现出高效性和准确性,并具备处理多条数据组合的潜在优化空间.该算法设计巧妙、复杂度低、易于实现,为COVID-19疫情数据监测与分析提供了新选择,也为未来类似决策问题的研究树立了典范.

A novel multi-variable decision tree algorithm based on decision class partitioning was utilized to construct decision logic for the COVID-19 Surveillance dataset,successfully deriving two equivalent and accurate decision expressions that classify the severity of the illness into mild,moderate,and severe categories,aligning with China's diagnostic and treatment protocols. The new algorithm outperforms two other algorithms in terms of time complexity,demonstrating efficiency and accuracy when processing a dataset with seven conditional attributes. It also possesses potential optimization space for handling combinations of multiple data entries. The algorithm is ingeniously designed,features low complexity,and is easy to implement,providing a new option for COVID-19 epidemic data surveillance and analysis and setting a paradigm for future research on similar decision-making problems.