COVID-19 Detection from Lung Tomography Images Using Deep Learning and Machine Learning Methods

EasyChair Preprint no. 4097

Abstract

Coronavirus (COVID-19) is an epidemic disease that spreads all over the world in a very short time and has fatal consequences. Such infectious diseases must be detected correctly without harming people or with minimal harm, and the necessary treatment must be initiated early. In this study, it was aimed to detect the COVID-19 image from different lung tomography images (COVID-19, viral pneumonia, bacterial pneumonia and normal) with artificial intelligence and machine learning techniques. In this context, K-Nearest Neighborhood (KNN) method, which is a machine learning algorithm, has been used together with Convolutional Neural Networks (CNN) and Deep Neural Networks (DNN), which are among the current techniques of artificial intelligence, deep learning approaches. In addition, in the CNN method, the results were tested by creating models with combinations of different optimization and activation functions and neuron numbers. Thus, the application potential of CNN, DNN and KNN methods in image recognition and classification was seen and the success of the proposed model was demonstrated with the findings obtained.

Keyphrases: CNN, COVID-19, Derin Öğrenme, DNN, KNN, Makine Öğrenmesi, Tomografi, Yapay Zekâ

@Booklet{EasyChair:4097,
  author = {Resul Bütüner and M. Hanefi Calp},
  title = {COVID-19 Detection from Lung Tomography Images Using Deep Learning and Machine Learning Methods},
  howpublished = {EasyChair Preprint no. 4097},

  year = {EasyChair, 2020}}