An efficient approach for classifying chest X-ray images using different embedder with different activation functions in CNN. Issue 2 (17th February 2021)

• Record Type: Journal Article
• Authors: Gupta, Amit; Gupta, Richa; Garg, Navin
• Is Part Of: Journal of interdisciplinary mathematics. Volume 24:Issue 2(2021)
• Page Start: 285
• Page End: 297
• Publication Date: 2021-02-17
• Subjects: 62M45 -- 92B20; Deep neural network -- COVID-19 -- CNN -- Inception V3 -- SqueezeNet -- ReLu -- Sigmoid -- Tanh; Mathematics -- Periodicals; Mathematics; Periodicals; 510.5
• DOI: 10.1080/09720502.2020.1838060
• Languages: English
• ISSNs: 0972-0502

Abstract

Abstract: The automated models play an important role to identify the diseases in the given chest X-ray images. This helps the doctors to speed up the diagnosis process so that the treatment of the patient will start as soon as possible. This paper shows the deep neural network model with different image embedding like SqueezeNet and Inception V3which uses four different activation functions and the accuracy of the output is over 99% of each model. The activation function increases the accuracy of the model so this paper give a comparison of the deep neural network which uses the different activation functions. This model is used to classify the chest X-ray images into binary classification of COVID-19 and non-COVID-19 images. The novel corona virus is a great jeopardy to the human life in 2019-20, thus earlier detection of the disease can slow down the spreading of the disease.