ABSTRACT
COVID-19 is a disease caused by the SARS-CoV-2 virus or often referred to as Corona Virus. In December 2019, this virus begin to spread from Wuhan, China to all over the world and was declared a pandemic. The virus attacks the respiratory tract so that sufferers have symptoms such as acute respiratory infection. In many cases, there are also patients with COVID-19 who do not have the following symptoms, making it difficult to determine the patient's COVID-19 status before a PCR test is performed. In this research, we try to do a rapid diagnosis with the final status of COVID-19 such as close contact, suspect, probable, and confirm, based on symptoms experienced by patients using the Adaptive Neuro-Fuzzy Inference System (ANFIS) method. ANFIS was chosen because ANFIS uses an artificial neural network concept that is suitable for use in patterned and complex calculations. ANFIS also has the basis of fuzzy logic that can map the expert and linguistic aspects of humans. Generated model from ANFIS training tested with entering symptoms patients data, then matched with COVID-19 status. Error calculation using MAE as an evaluation of the accuracy of this model. Evaluation is based on 10-fold cross validation. The experimental results obtained an accuracy of 82.39% with an MAE value of 0.1558 for training and 0.1903 for testing. © 2022 IEEE.
ABSTRACT
The Covid-19 outbreak, which has been declared a pandemic since March 2020, has been causing problems worldwide. As a result, many countries have implemented lockdown policies to control the spread of the Covid-19 virus. In addition, time spent on gaming activity has increased by 52% since video game engagement was thought to be essential in improving players' vitality, reducing psychological suffering, and helping combat stress. This literature review was conducted as a systematic literature review based on the 15 primary studies between 2020 and March 2022. Analysis of the selected primary studies revealed that the authors conducted studies of gaming activities in the Covid-19 pandemic era for four reasons: To determine the factors of play intention, factors of purchase intention, factors of gaming disorder, and to investigate the impact of the gaming activities itself. Physical health issues, family interactions, social interactions, fear of missing out, psychological distress, and time and location flexibility are the six determinants for people to continue to play video games. Meanwhile, the expectation of performance and effort were the factors that influenced purchase intention in mobile games. And from these determinant factors, it seemed that psychological distress and fear of missing out were the causes of someone experiencing a gaming disorder. As for the method employed, Structural Equation Modeling (SEM) was the most extensively used statistical tool in conducting quantitative research. Six of the eleven quantitative research in the primary studies utilized SEM, and the others employed other statistical tools. Although, in the selected primary studies, we also have four studies conducted qualitative research using interviews and open-ended surveys. Ten different countries were identified as the origin country of the respondents for the primary studies, with Finland and the United States as the most research object. However, we also found three studies that did not specifically mention the origin countries of the respondents. © 2022 IEEE.
ABSTRACT
Coronary heart disease (CHD) is a comorbidity of COVID-19;therefore, routine early diagnosis is crucial. A large number of examination attributes in the context of diagnosing CHD is a distinct obstacle during the pandemic when the number of health service users is significant. The development of a precise machine learning model for diagnosis with a minimum number of examination attributes can allow examinations and healthcare actions to be undertaken quickly. This study proposes a CHD diagnosis model based on feature selection, data balancing, and ensemble-based classification methods. In the feature selection stage, a hybrid SVM-GA combined with fast correlation-based filter (FCBF) is used. The proposed system achieved an accuracy of 94.60% and area under the curve (AUC) of 97.5% when tested on the z-Alizadeh Sani dataset and used only 8 of 54 inspection attributes. In terms of performance, the proposed model can be placed in the very good category. © 2022 The Korea Institute of Information and Communication Engineering. All Rights Reserved.