Downloads: 1 | Views: 103 | Weekly Hits: ⮙1 | Monthly Hits: ⮙1
Research Paper | Computer Engineering | Iraq | Volume 12 Issue 6, June 2023
Enhanced Machine Learning Model for Prediction of COVID-19 Cases in Iraq
Zakarya A Mohamed Zaki [2] | Aisha Hassan Abdalla [2]
Abstract: The SARS-CoV-2 virus is responsible for the emergence of the highly contagious illness known as COVID-19. The disease has been classified as a global pandemic, impacting millions of people throughout the globe. It has created a change in the research community's orientations for identification, analysis, and control via the application of different statistical and predictive modelling methodologies. These numerical models are examples of decision-making techniques that depend significantly on data mining and machine learning to create predictions based on historical data. In order to make smart judgments and create strong strategies, policymakers and medical authorities need reliable forecasting techniques. These studies are carried out on a variety of small scale datasets including a few hundreds to thousands of records. This study uses a large dataset consisting of COVID-19 instances recorded on a daily basis in Iraq, together with socio-demographic and health related attributes for the region. The primary goal of the research is to do daily forecasting of Covid-19 instances using time series forecasting. The predictive modeling for daily COVID-19 infection cases involved several neural network architectures, including artificial neural networks, convolutional neural networks (CNNs), long short-term memory networks (LSTMs), and Hybrid CNN-LSTM model. Prior to the modeling, appropriate procedures were used to prepare the data and identify any seasonality, residuals, and trends. The contribution of this work lies in the development of an enhanced forecasting model for COVID-19 infection cases. It utilizes a combination of different neural network models to create an effective forecasting tool. The proposed enhanced hybrid model built using a Convolutional Neural Network and a Long Short-Term Memory network (EH-CNN-LSTM). The model is trained and tested on various subsets of the dataset. It is discovered that the higher the amount of training data, the better the predicted performance. Compared to other models, the proposed EH-CNN-LSTM performs better. Mean Absolute Percentage Error (MAPE), Mean Squared Logarithmic Error (MSLE), and Root Mean Squared Logarithmic Error (RMSLE) are used to evaluate the predictive performance. EH-CNN-LSTM, which was trained on 80% of the data and evaluated on 20% of the data, achieved a MAPE of 5.28, MSLE of 0.00, and RMSLE of 0.02.
Keywords: COVID-19, Prediction, Forecasting, and Machine Learning
Edition: Volume 12 Issue 6, June 2023,
Pages: 2249 - 2260