خلاصة:
Corona Virus (COVID-19) is a major problem among people, and it causes suffering worldwide. Yet, the traditional prediction models are not yet suitably efficient in catching the fundamental expertise as they cannot visualize the difficulty in the health's representation problem areas. This paper states prediction mechanism that uses a model of deep learning called Long Short-Term Memory (LSTM). We have carried this model out on corona virus dataset that obtained from the records of infections, deaths, and recovery cases across the world. Furthermore, producing a dataset which includes features of geographic regions (temperature and humidity) that have experienced severe virus outbreaks, risk factors, spatio-temporal analysis, and social behavior of people, a predictive model can be developed for areas where the virus is likely to spread. However, the outcomes of this study are justifiable to alert the authorities and the people to take precautions.
ملخص الجهاز:
Long Short-Term Memory Approach for Coronavirus Disease Prediction Omar Ibrahim Obaid / Department of Computer Science, College of Education, AL-Iraqia University, Baghdad, Iraq.
This paper states prediction mechanism that uses a model of deep learning called Long Short-Term Memory (LSTM).
We have carried this model out on corona virus dataset that obtained from the records of infections, deaths, and recovery cases across the world.
Furthermore, producing a dataset which includes features of geographic regions (temperature and humidity) that have experienced severe virus outbreaks, risk factors, spatio-temporal analysis, and social behavior of people, a predictive model can be developed for areas where the virus is likely to spread.
Keywords: Deep learning, LSTM, Prediction, COVID-19, Recurrent Neural Network (RNN).
Nowadays, number of coronavirus (COVID-19) infections were 428,405, and the deaths was over 19,000 at the time of writing this research paper in April according to the World Health Organization (WHO.
In the last decade, many researchers around the world have applied deep learning approaches to predict illness based on the datasets in the medicine field.
Due to (RNNs) have two major issues: a gradient disappearing and the gradient crashing, that most make it useless, long short-term memory (LSTM) approach is solved this problem, and we use it in this paper as it is most suitable technique for time series problems.
Deep Learning for Computer Vision: Expert techniques to train advanced neural networks using TensorFlow and Keras / Bibliographic information of this paper for citing: Ibrahim Obaid, Omar; Abed Mohammed, Mazin & Mostafa, Salama A.