Page 29 - AKSES vol3
P. 29
AKSES AR TIKEL PEN Y ELIDIK AN ADV ANCING KNOWLEDGE FOR SUC CESS FT SM UKM
LIVING ANALYTICS: LEVERAGING SOCIO ECONOMIC DATA FOR Six new income classes are defined, namely B40, M40, T20, B40-FB (B40 class with Financial Burden Risk), M40-FB (M40
THE NATION WELLBEING class with Financial Burden Risk), and T20-FB (T20 class with Financial Burden Risk. (accuracy 95%)
CASE 3: Multidimensional Poverty Indicators For
Azuraliza Abu Bakar, Zulaiha Ali Othman, Zalinda Othman, Suhaila Zainudin, Nor Samsiah Sani, Rusnita Hamdan The Bottom 40 Percent Group
azur aliza @uk m.edu .m y
The case aims to identify the right indicators and dimensions that will provide data-driven MPI measurement. It uses
LIVING ANALYTICS combines analytical research and development in areas involving the people and national well- the Malaysian census data, consisting of 532,298 households, and employs K-Means clustering approach to identify the
being, to enhance the high-impact society in Malaysia. It is a data-intensive technology that enables the Government multidimensional poverty Indicators for the bottom 40 percent group. The analysis discovered seven multidimensional
and public to get new solutions and initiatives according to personal needs, profiles, and activities by collaborating with poverty indicators from three dimensions encompassing education, living standard, and employment. Out of the seven
the public sector to conduct an analytical study of data by fully utilizing analytic advantages (adopted from LARC 2017). indicators, this study proposed six indicators to be added to the current MPI to establish a more meaningful scenario
It aligns very well with the Government’s vision to make high-impact people and society. It ensures all citizens can enjoy of the current poverty trend in Malaysia. The model may help the government adequately identify the B40 group who
a better quality of life, where two of the pillars are suffers from financial burden, which could have been currently misclassified.
CASE 4: Gendering Analytics
1. To strengthen inclusivity towards equitable society and
Gendering analytics aims to determine gender-based factors that contribute to the financial risk or non-risk of a household.
2. To improve the well-being of the people.
It employs machine learning towards the Census dataset to find households with financial risk patterns from different
states. The patterns obtained showed that women are more economically vulnerable, and most of them only achieve
Poverty is a barrier for economies to harness the opportunities of the fourth industrial revolution effectively. Data analytics
integrate socio-economics data from several sources and employ machine learning to discover important indicators and financial stability after the age of 30. In addition, the tree reveals that non-Bumiputra women are more financially vulnerable
patterns that contribute to effective government decision-making, policies, initiatives, and actions. compared to their Malay counterparts.
CASE 5: Households Overspending Model of B40,
Socio-economic data involves data related to the people such as census survey data, welfare and poverty data, M40, And T20 Income Class
demographics, housing, transportation, amenities and infrastructure, economics, retailing, health and medical, This case aims to find important factors that affect the spending patterns among households. It develops a household
and education. These data can be obtained from field surveys, street interviews, government statistics on the entire overspending model by using machine learning towards 4 million household expenditure records obtained from the
population. It can also form government administrative records, e.g., tax records, auto registrations, and property taxes. survey conducted in 2016 by the Department of Statistics Malaysia. The model is developed using 12 demographic
In Malaysia, among the socio-economics data are the Household Expenditure and Income Survey from Department attributes with 14451 records. The model showed that the six attributes that influence most to overspending are state,
of Statistic Malaysia (3,793,433 instances and 31 attributes), E-Kasih Dataset (196,650 observations and 24 variables) – race, income, strata, number of households, and categories. The model proved that the number of household members
Poverty data bank Malaysia – Economic Planning Unit, Prime Minister Department, Bantuan Rakyat 1Malaysia (BRIM) could be one of the variables in identifying the poverty category as B40, M40, or T20.
dataset.
CASE 6: Financial Literacy Model of Income Class
Several baselines can be used in the socio-economics analytics, such as published reports and domain experts, most In Malaysia
notably from Khazanah Research Institute, Department of Statistics Malaysia, and the Economic Planning Unit. Example This case study finds the factors that contribute to the risk of the financial literacy of a household using a Malaysian
baselines are the Household income baseline, Classification Of Individual Consumption According To Purpose (COICOP), household and expenditure dataset. The machine learning model classifies the level of financial literacy among the
Global Indicator Of Multidimensional Poverty Index (MPI), and the Multidimensional Poverty Index (MPI). The following household income class.
are six use cases that employed Machine Learning algorithms for socio-economics data analytic model development.
Big data research would bring forth novel data, methods, and evaluation challenges to provide insightful data findings in
C A S E 1 : Multidimensional Poverty Classification Model
socioeconomics. The large amount of data obtained from diverse sources could present significant challenges when one
The Multidimensional Poverty Classification uses the eKasih 2017 dataset with 196,650 obs. of 15 variables and employed attempts to connect and correlate multiple data linkages. Data Analytics could improve Government policy, inter-agency
the Random Forest algorithm to classify poverty into two classes: poor and hardcore poor. The model also produces coordination, and efficient government aid coordination, target based initiatives, and programs. The research may
multidimensional poverty indicators. generate new measures; initiatives that lead to policy improvements are essential to ensure the most critical indicator
besides income. The emergence of the 4th Industrial Revolution (4IR) has a significant impact on the economy, society,
C A S E 2 : Financial Burden Risk Prediction Model and politics. Technological progress in 4IR is a primary driver of aggregate economic growth and living standards over
the long term. It increases overall productivity, thereby boosting per capita income and consumption, affecting the low-
The financial burden risk prediction model uses Households Expenditure and Income Dataset 2014-2016 to find the income group directly.
factors that contribute to the risk of the financial burden and the patterns of households with the financial burden from
different states.