Experimental Analysis Using Hybrid Convolutional Neural Networks, Gradient Boosting Classifier, and Differential Algorithm for Detection of COVID-19 from X-Ray Images

A significant number of individuals have lost their lives due to the new COVID-19 virus. The coronavirus has ruined
many people’s lives, and the healthcare system is struggling a lot because of it. Since the virus can harm the lungs severely, it’s
essential to find it early. To detect COVID-19 from X-ray images, this study presents a novel hybrid approach that combines
convolutional neural networks, gradient-boosting classifiers, and differential algorithms. This strategy offers a synergistic
fusion of deep learning, ensemble learning, and optimization strategies. In the context of COVID-19 detection by X-ray imaging,
the adaptive integration of these disparate methodologies constitutes a groundbreaking attempt to address the shortcomings of
current methods and significantly improve diagnostic accuracy. This study recommends using a computer program called CNN
to help identify COVID-19 in chest X-ray images. For this study, scientists used a collection of 13,000 chest X-ray pictures. With
CLAHE’s help, researchers improved the original dataset. The research used advanced computer programs to find essential
details in the pictures and then used a method to focus on the most valuable parts. To prevent overfitting, the model locks in the
weights of the dense layers trained in previous rounds. This enables it to fit the new thick layer and optimize the convolutional
layers while retaining the previously learned data. The final layer of the CNN Model was replaced with the Gradient Boosting
Machines classifier for classification. The results showed that the suggested approach was 98% specific, 97% sensitive, and
98% accurate. According to study data, the suggested approach performed better than previous COVID-19 detection
investigations based on X-ray imaging