Aryatama Fonggi Yudi, Kurniadi Felix Indra, Manik Ngarap Im
Mathematic Department, School of Computer Science, Bina Nusantara University, Jakarta, 11530, Indonesia.
Computer Science Department, School of Computer Science, Bina Nusantara University, Jakarta, 11530, Indonesia.
Procedia Comput Sci. 2023;216:120-127. doi: 10.1016/j.procs.2022.12.118. Epub 2023 Jan 10.
COVID-19 was originally diagnosed in Wuhan, China, in late December 2019; it subsequently expanded internationally, affecting around 7 million people and caused 300,000 deaths by May 2020. An Application that could surveillance the Covid-19 in Indonesia is needed. Therefore, we proposed a prediction application for the COVID-19 pandemic situation in Indonesia by referring to public compliance with surveillance policies using the SPCIRD model. The Levenberg Marquardt curve fitting method was chosen because it is simple and produces a reasonably good model. According the result of questionnaire and black box testing our application performed really well. The result of our SPCIRD model with Levenberg-Marquardt optimization achieved R2 with the score -1.248 in active cases, -0.235 in recovered and -3.982 in death for 281 number of iterations. We are also achieved 93.3% that agree and very agree with the application usability to understand and to predict pattern from COVID-19. It was also had 83.3% respondent that satisfy with the prototype.
2019年12月下旬,新型冠状病毒肺炎(COVID-19)最初在中国武汉被诊断出来;随后它在国际上蔓延,到2020年5月,感染人数约达700万,造成30万人死亡。因此,需要一个能够监测印度尼西亚COVID-19情况的应用程序。为此,我们参考公众对监测政策的遵守情况,使用SPCIRD模型,提出了一个针对印度尼西亚COVID-19疫情形势的预测应用程序。选择Levenberg Marquardt曲线拟合方法是因为它简单且能产生相当不错的模型。根据问卷调查和黑盒测试的结果,我们的应用程序表现良好。我们采用Levenberg-Marquardt优化的SPCIRD模型在281次迭代中,活跃病例的R2得分为-1.248,康复病例为-0.235,死亡病例为-3.982。我们还发现,93.3%的人同意并非常同意该应用程序在理解和预测COVID-19模式方面的可用性。此外,83.3%的受访者对该原型感到满意。