Patient triage at emergency departments (EDs) is necessary to prioritize care for patients with critical and time-sensitive conditions. In this paper, the metaheuristic optimization algorithms simulated annealing (SA) and adaptive simulated annealing (ASA) are proposed to optimize the parameters of extreme gradient boosting (XGB) and categorical boosting (CaB). The proposed algorithms are SA-XGB, ASA-XGB, SA-CaB, and ASA-CaB. Grid search (GS), a traditional approach used for machine learning fine-tuning, is also used to fine-tune the parameters of XGB and CaB, which are named GS-XGB and GS-CaB. The optimized model is used to develop an e-triage tool that can be used at EDs to predict ED patients' emergency severity index (ESI). The results show ASA-CaB outperformed all the proposed algorithms with accuracy, precision, recall, and f1 of 83.3%, 83.2%, 83.3%, and 83.2%, respectively.

急诊科 (ED) 需要对患者进行分类，以便优先护理病情危重且时间敏感的患者。本文提出了元启发式优化算法模拟退火（SA）和自适应模拟退火（ASA）来优化极限梯度提升（XGB）和分类提升（CaB）的参数。提出的算法是 SA-XGB、ASA-XGB、SA-CaB 和 ASA-CaB。网格搜索（GS）是一种用于机器学习微调的传统方法，也用于微调XGB和CaB的参数，分别命名为GS-XGB和GS-CaB。优化后的模型用于开发电子分诊工具，可在急诊室用于预测急诊室患者的紧急严重程度指数 (ESI)。结果表明 ASA-CaB 优于所有提出的算法，准确率、精确度、召回率和 f1 为 83.3%，