Journals Information
Universal Journal of Applied Mathematics Vol. 12(4), pp. 85 - 94
DOI: 10.13189/ujam.2024.120401
Reprint (PDF) (1099Kb)
Improvisation of Fuzzy C-Least Median Using Canberra Distance onto Multiple Linear Regression towards Malaysian Household Income
Anis Nelissa Abdul Hamid , Shahirulliza Shamsul Ambia , Sumarni Abu Baka *, Noratika Nordin
College of Computing, Informatics and Mathematics, Universiti Teknologi MARA (UiTM), Malaysia
ABSTRACT
This manuscript introduces a novel enhancement to the Fuzzy C-Least Median (FCLM) model by incorporating Canberra distance into Multiple Linear Regression (MLR) frameworks, aiming to address the intricate challenge of predicting ambiguity in data clustering. The study examines household income and demographic data in Malaysia and systematically assesses the performance of the MLR model alongside the proposed FCLM-CD MLR model. The evaluation includes comparisons with MLR, FCM MLR (Fuzzy C-Means Multiple Linear Regression), and FCLM MLR (Fuzzy C-Least Median Multiple Linear Regression) models. The FCLM-CD MLR model incorporates the Fuzzy C-Least Median (FCLM) algorithm, enriched with Canberra distance, to address clustering ambiguity. Key evaluation metrics such as Root Mean Square Error (RMSE) and R-Squared values are employed to assess predictive accuracy and explanatory power. Results reveal that the FCLM-CD MLR model outperforms the standard MLR model, FCM-MLR and FCLM-MLR, demonstrating superior predictive accuracy and enhanced explanatory power. The integration of FCLM with Canberra distance represents a significant methodological advancement, offering a promising approach to addressing clustering challenges and capturing variations in household income effectively. Notably, the incorporation of Canberra distance into FCLM emerges as a pivotal element in bolstering the effectiveness of predictive modelling. This study contributes to the field by overcoming clustering challenges and providing nuanced insights into the complexities of household income and demographic data. The amalgamation of FCLM and Canberra distance presents a methodological innovation with broad implications for predictive modelling across diverse domains, enriching the landscape of data analysis methodologies.
KEYWORDS
Canberra, Clustering, Fuzzy C-Least Median, Multiple Linear Regression
Cite This Paper in IEEE or APA Citation Styles
(a). IEEE Format:
[1] Anis Nelissa Abdul Hamid , Shahirulliza Shamsul Ambia , Sumarni Abu Baka , Noratika Nordin , "Improvisation of Fuzzy C-Least Median Using Canberra Distance onto Multiple Linear Regression towards Malaysian Household Income," Universal Journal of Applied Mathematics, Vol. 12, No. 4, pp. 85 - 94, 2024. DOI: 10.13189/ujam.2024.120401.
(b). APA Format:
Anis Nelissa Abdul Hamid , Shahirulliza Shamsul Ambia , Sumarni Abu Baka , Noratika Nordin (2024). Improvisation of Fuzzy C-Least Median Using Canberra Distance onto Multiple Linear Regression towards Malaysian Household Income. Universal Journal of Applied Mathematics, 12(4), 85 - 94. DOI: 10.13189/ujam.2024.120401.