Machine Learning And Coronavirus Disease Prediction

Main Article Content

Yusmi Mohd Yunus, Muhammad Hisyam Zakaria, Mohd Fazril Izhar Mohd Idris, Na Izni, Mohd Norazmi Nordin


This Study Is Done By Using The Majority Rule At The End Of The Process I.E. Covid19 Detection Is Positive When More Than Half Of The Architectures Points In Favour Of It Otherwise Negative. Overall, Results Obtained Demonstrate A Strong Effect Of Deep Learning Architectures On The Covid19 X-Ray Datasets. We Also Discussed Recently Available Datasets Around The Globe And The Application Of Various Dl Architectures. Lenet5, Cnn, Dense-Net121, Densenet169, Densenet201, Resnet50, Vgg16, Vgg19, Mobilenetv2, Nasnetmobile, Nasnetlarge, Inceptionv3, Inceptionresnetv2 And Xception Were Presented With Performance Measures As A Proof Of Concept. Further, We Proposed A Method To Detect The Covid-19 Presence Based On The Results Of The Above Architectures. X-Ray Diagnosing Can Be Used As An Initial Method During Large Population Testing And Can Be Made Easily Available At Any Remote Place With Good Internet Connection. Future Studies Can Include Adding More Data But Not Limited To X-Ray Images. Moreover, Covid-19 Diagnosing With Sonography (Lung Ultrasound) Combined With Radiography Can Be Used To Increase The Detection Power As Ultrasound Frequency Analysis Using Acoustic Models Would Be Good Enough In Identifying Covid-19 Presence.


Article Details