Identifying and Predicting Muslim Community Funeral Funding Protocols

Abstract

This research aims to understand funeral poverty among the Muslim community in Malaysia by using Machine Learning algorithms. Generally, funeral poverty in Malaysia can be described as funerals in Malaysia being exorbitantly priced, inflicting a disproportionate amount of financial hardship on the poor. As the death rate continues to climb year after year, we may conclude that funeral fees are an unavoidable risk. This research problem was inspired by Sustainable Development Goal number 1 and 3 where the goal of this initiative is to increase awareness about the need of having funeral expenditures and how to prevent funeral poverty. Previous works have shown promising results. However, they are not conclusive enough as they lack predicting capability that Machine Learning can offer. Selected Machine Learning algorithms such as Decision Tree, Random Forest, and Naïve Bayes were used to classify the people that will go through funeral poverty based on a selected dataset and a survey conducted. The research methodology is as follows: 1. Collecting the data, 2. Pre-Process the data, 3. Exploratory data analysis, 4. Selecting the feature, 5. Modeling, 6. Evaluation. The results showed that the accuracy value for the selected algorithm is 0.95, 0.95, and 0.95 for the awareness and 0.8, 0.8, and 0.75 for B40 Status in survey datasets. With a larger taxonomy and more extensive, diverse sets of data, these figures are expected to improve. Linear Regression has been applied to help understand the features of a death rate. Both outcomes serve distinct reasons. Because the individuals who would experience funeral poverty are in lower-income neighborhoods, it was necessary to classify impoverished communities in order to anticipate funeral poverty. Following that, the findings of what causes death to rise aid in understanding how the B40 community has been impacted by funeral poverty. In the future, this study will aid future research on funeral poverty on a larger scale. When the two datasets are integrated rather than separated, better results can be obtained. The accuracy of the findings may also be improved by employing a wider range of Machine Learning methods and deep learning implementation. A more coherent and detailed model for forecasting funeral poverty can be developed.