Prediction of probability of death of COVID-19: Malaysian case study

Mohamed, Arina (2022) Prediction of probability of death of COVID-19: Malaysian case study. [Project Paper] (Submitted)

Text FS 2022 15.pdf Download (2MB)

Abstract

The world was shocked when the first series of COVID-19 cases were reported in late December 2019 in Wuhan, China. Clusters of infections has spread outside China which later urged the World Health Organization (WHO) to call it a pandemic. The pandemic has not only burdened the healthcare sectors, but it has also caused the economy to crippled causing millions of people struggling to carry on with their life. When an outbreak of new and unknown disease happens, it is very important to study the factors of the infections. Also, it is very crucial to investigate the trend and predict the upcoming situation so that the government can take a better plan to lessen the risk of the infections. This project aims to compare the result based on different prediction input parameters, which are Malaysian state, patients’ age and medical history. Also, this project intent to investigate the relationship between the number of death with number of cases from yesterday’s up to 10 days. This lagging of days is called as lags. To achieve the objectives, a set of new confirmed cases data and death cases data for each state, patient’s age and medical history are collected and analyzed before being divided into two parts for training and testing. This project used machine learning model, ARIMAX to predict the probability of death for each of the state, patient’s age and medical history. From the findings, the correlation across the data set discovered that the data set of new confirmed cases correlates with the data of death cases for each state, patient’s age and medical history with average of 0.6875. The accuracy and errors for each of the death cases data for state, patients’ age and medical history differ depends on the size of the training and testing data. For prediction of death in each state, W.P Labuan has recorded the lowest error, Root Mean Square Error (RMSE) and Mean Average Error (MAE) of 1.8696 and 1.4606 respectively. While the prediction of death for each of the patient’s age, the class age of 25-29 has recorded the lowest value of RMSE which is 1.0718. Lastly, Dyslipidemia has recorded the lowest RMSE and MAE, 3.0629 and 2.207 respectively compared to other five diseases of the patients’ medical history. Overall, the prediction of death for each disease history recorded the least average error compared to the data of each state and patient’s age. Also, the trend of lags in 10 days differ depends on the size of training and testing data.

Item Type:	Project Paper
Faculty:	Fakulti Sains
Depositing User:	Ms. ROHANA ALIAS
Date Deposited:	25 Jun 2024 10:08
Last Modified:	25 Jun 2024 10:08
URI:	http://psaspb.upm.edu.my/id/eprint/1981

Actions (login required)

View Item