The outbreak of the coronavirus is in its growing stage due to the lack of standard diagnosis for the patients. The situation of any populous area in a geographic location is very critical due to the quick virus spread from an infected individual to the rest. Currently, medical administration is at a crisis point due to the rapidly increasing number of cases and limited medical facilities. Thus, it is time to explore and design an intelligent model to monitor patient health symptoms remotely and predict and detect the abnormality of the patient's health status in quick succession. Thus, the health status of a coronavirus-affected patient can be identified via a well-adjusted predictive model by analyzing the observed parameters of the health. In the proposed model, an Auto-regressive Integrated Moving Average is incorporated to design a predictive model to find the kth forecast of the observed health symptoms of a patient, and Akaike Information Criteria based selection is introduced to find the current best-fit prediction model. Further, the features are extracted from the forecast over each symptom to find a pattern of each patient, and the patterns are learned by the K-Means algorithm to detect the symptomatic and asymptomatic patient intelligently. To demonstrate the efficiency of the proposed model, we evaluate the model using a synthetic dataset, generated from the health symptoms of 400 patients and compare the performance of the model with the standard methods.

中文翻译：

---

用于检测有症状/无症状 COVID-19 患者的智能健康监测系统


由于缺乏对患者的标准诊断，冠状病毒的爆发正处于增长阶段。由于病毒从感染者到其他人的快速传播，某个地理位置上任何人口稠密地区的情况都非常严峻。当前，由于病例数迅速增加、医疗设施有限，医疗管理正处于危机时刻。因此，需要探索和设计一种智能模型来远程监测患者健康症状，并快速连续预测和检测患者健康状况的异常。因此，可以通过分析观察到的健康参数，通过经过良好调整的预测模型来确定受冠状病毒影响的患者的健康状况。在所提出的模型中，结合了自回归积分移动平均线来设计预测模型，以找到患者观察到的健康症状的第 k 个预测，并引入基于 Akaike 信息标准的选择来找到当前的最佳拟合预测模型。此外，从每个症状的预测中提取特征，找到每个患者的模式，并通过K-Means算法学习该模式，以智能地检测有症状和无症状的患者。为了证明所提出模型的效率，我们使用根据 400 名患者的健康症状生成的合成数据集来评估该模型，并将模型的性能与标准方法进行比较。