Application of fuzzy linear regression models for predicting tumor size of colorectal cancer in Malaysia's Hospital

• Main Author: Shafi, Muhammad Ammar
• Format: Thesis
• Language: English; English
• Published: 2015
• Subjects: QA273-280 Probabilities. Mathematical statistics
• Online Access: http://eprints.uthm.edu.my/1644/1/24p%20MUHAMMAD%20AMMAR%20SHAFI.pdf; http://eprints.uthm.edu.my/1644/2/MUHAMMAD%20AMMAR%20SHAFI%20WATERMARK.pdf; http://eprints.uthm.edu.my/1644/

Fuzzy linear regression analysis has become popular among researchers and standard model in analysing data vagueness phenomena. These models were represented by five statistical models such as multiple linear regression, fuzzy linear regression (Tanaka), fuzzy linear regression (Ni), extended fuzzy linear regression under benchmarking model (Chung) and fuzzy linear regression with symmetric parameter (Zolfaghari). A case study in colorectal cancer (CRC) data at the general hospital in Kuala Lumpur was carried out using the five models as mention above. Secondary data of 180 colorectal cancer patients who received treatment in general hospital were recorded by nurses and doctors. Twenty five independent variables with different combination of variable types were considered to find the best models to predict the size of tumor colorectal cancer. The quality of life among CRC patients which is to detect the early CRC stage is still very poor, not implemented and divulged as a nationwide programme. The main objective of this study is to determine the best model by predicting the size of tumor of CRC. Moreover, this study wants to identify the factors and symptoms that contribute the size of tumor. The comparisons among the five models were carried out to find the best model by using statistical measurements of mean square error (MSE) and root mean square error (RMSE). The results showed that the fuzzy linear regression with symmetric parameter (Zolfaghari) was found to be the best model, having the lowest MSE and RMSE value by 98.21 and 9.91. Hence, the size of tumor could be predicted by managing twenty five independent variables.