Prediction for high-risk symptoms of lung cancer in Malaysia using fuzzy linear regression model

• Main Author: Zakaria, Aliya Syaffa
• Format: Thesis
• Language: English; English; English
• Published: 2023
• Subjects: R858-859.7 Computer applications to medicine. Medical informatics
• Online Access: http://eprints.uthm.edu.my/10971/1/24p%20ALIYA%20SYAFFA%20ZAKARIA.pdf; http://eprints.uthm.edu.my/10971/2/ALIYA%20SYAFFA%20ZAKARIA%20COPYRIGHT%20DECLARATION.pdf; http://eprints.uthm.edu.my/10971/3/ALIYA%20SYAFFA%20ZAKARIA%20WATERMARK.pdf; http://eprints.uthm.edu.my/10971/

Lung cancer has been recorded as the most common cancer globally, contributing 12.2% of all new cases diagnosed in 2020, with the greatest mortality rate due to its late diagnosis and poor symptom detection. Nowadays, Malaysia has reached 4,319 lung cancer deaths, accounting for 2.57 per cent of all deaths in 2020. Late diagnosis is the norm for lung cancer, which makes survival challenging and the likelihood of recovery low. Nevertheless, in Malaysia, most cases are discovered late, when the tumors have grown too far, or the disease has spread to other body parts that cannot be removed through surgery. This situation frequently occurs due to the lack of public knowledge among Malaysians regarding cancer-related signs and symptoms. Therefore, Malaysians must be aware of the high-risk symptoms of lung cancer to increase the survival rate and decrease the mortality rate. This study aims to compare multiple linear regression and fuzzy linear regression model using a triangular fuzzy number proposed by Tanaka. The H-value from 0.0 to 1.0 is adjusted to find the optimal value of an objective function to predict high-risk lung cancer symptoms in Malaysia. The secondary data is analyzed using the fuzzy linear regression model, which can reduce the interference of irrelevant information and improve the precision of the results. This research data was collected from patients with lung cancer at Al-Sultan Abdullah Hospital (UiTM Hospital), Selangor. The data of 124 lung cancer patients were analyzed using Microsoft Excel and MATLAB. The study implemented measurement error of cross-validation technique, which is mean square error (MSE) and root mean square error (RMSE), to enhance data accuracy. The results show that haemoptysis and chest pain has been proven to be the highest risk, among other symptoms acquired from the data analysis. It has been determined that H-value of 0.0 has the smallest measurement error, with MSE of 1.455 and RMSE of 1.206 as the multiple linear regression method has the MSE value of 306.257 while the RMSE has the value of 17.500