The COVID-19 epidemic has given an unusually terrible circumstance for the entire planet, terrifyingly stopping life, and taking thousands of lives. COVID-19 continues to pose a serious threat to the public health system due to its expansion to 212 nations and territories, rising incidence of infection cases, and death tolls of 5,212,172 and 334,915 (as of May 22, 2020). This initiative uses artificial intelligence (AI) to produce a response to fight the infection. Several deep learning (DL) techniques have been demonstrated to accomplish this, including long/short term memory, extreme learning machine (ELM), and generative adversarial networks (GANs). It describes an integrated approach to bioinformatics that combines a number of features for both structured and unstructured data sources to produce user-friendly platforms for physicians and academics. The speedier detection and treatment of COVID-19 illnesses is the main advantage of these AI-based solutions. In an attempt to identify the approach with the best accuracy and results, the work proposed a number of different approaches.
Top1. Introduction
From the common cold to more serious illnesses like severe acute respiratory syndrome (SARS), a coronavirus is a type of virus that causes respiratory infections. The corona virus's structure is depicted in Figure 1. 2020 saw the global emergence of SARS-CoV-2, a novel coronavirus. COVID-19 is a disease that is caused by SARS-CoV-2 and has spread over the world. The coronavirus has recently spread quickly as a new disease throughout the world. Typically, humans or animals are the sources of coronavirus transmission. According to the animal transmission, bats are the primary means of transmission.
Additionally, researchers have found that a variety of common coronaviruses, including 229E: Alpha, NL63: Alpha, OC43: Beta, and HKU1: Beta, can spread within the human body. Other human beta coronaviruses that produce diseases comparable to MERS, SARS, or SARS-COV are also present in the environment. These include SARS-COV-2, the new coronavirus that causes COVID-19, and MERS-COV (Coronavirus, 2020). Most experts or researchers have highlighted the unique COVID-19 diagnostic as shown in Figure 1.
1.1 Causes of COVID-19
Respiratory droplets are how SARS-CoV-2 spreads. An uninfected person's mouth, eyes, or nose may get infected droplets when an infected person coughs, sneezes, or talks. This is the most typical way that COVID-19 distributes. Less frequently, touching an infected surface and then touching one's lips, nose, or eyes can transfer COVID-19. Symptoms of COVID-19 might appear anywhere between two and fourteen days following exposure to SARS-CoV-2. They resemble the signs of a cold or the flu and might consist of:
CNNs are the most potent deep learning method because they can be trained on enormous amounts of data. However, in a few instances, it was unable to match the demand and simply provides a simple yes/no answer as to whether the patient has an infection. Therefore, CNN models follow precise procedures to train such datasets on important data that will be used to explain the final predictions. The many areas of a patient's lungs affected by a coronavirus (COVID-19) have been identified using a single straightforward example.
In this project two activation function are used one is ReLU and second one is Sigmoid. In a neural network, the activation function is responsible for translating the total weighted input of the node into the activation or output of the node for that input.