A deep learning framework for recognizing COVID-19 positive people was developed using an analysis of cough noises. There has been a COVID-19 pandemic which has a major influence on the world's health system, leading to 231 million illnesses and 4.7 million fatalities. By combining a pre-trained embedding and manually created features taken from audio recordings of draw sounds, a feature representation for cough noises is created. The front-end feature extraction procedure is known as this. Next, various back-end classification models are fed the aggregated characteristics to recognize COVID-19 positive patients. Using six deep learning in COVID-19 prediction has important ramifications for the healthcare industry. An extraordinary worldwide health catastrophe caused by the pandemic of COVID-19 has highlighted the requirement for novel, quick diagnostic techniques. In order to develop a quick and effective diagnostic tool, this research investigates the using deep learning algorithms. Regarding the analysis of cough in the COVID-19 identification. The collection of data is ready to get a deep learning model trained using techniques for preprocessing data, such as noise reduction and feature extraction. Convolutional Neural Networks (CNNs) and recurrent neural networks are two examples of deep learning architectures.