International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 03 | Mar 2025
p-ISSN: 2395-0072
www.irjet.net
Global Health Expenditure Analysis and Predictions Nenturi Vedha Sri1, Dharavath Vandhana2, Koukuntla Sneha3, Dr M V Krishna Rao4 1,2,3B. Tech Student, Dept. of Computer Science and Engineering (Data Science), Institute of Aeronautical
Engineering, Telangana, India
4Professor, Dept. of Computer Science and Engineering (Data Science), Institute of Aeronautical Engineering,
Telangana, India ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - This study analyzes the trends of global health
Over the last two decades, the landscape of global health has witnessed multiple transformations expenditure due to economic dynamics, population aging, technological advancements, inflation in healthcare, and public health crises like COVID-19. These changes have warranted the need for further analytical techniques to uncover expenditure trends, advocate financial sustainability, and assess future scenarios. For government officials, researchers, and health institutions, such knowledge provides a basis to enhance resource allocations, improve financial planning, and sustain health systems in the future. Data is collected from the World Health Organization (WHO) Global Health Expenditure Database, providing data for nearly 200 countries between 2000 and 2022.This dataset covers all aspects related to government and private contributions toward health, per capita expenditure, and other economic trends concerning the global healthcare systems. The size and intricacy of the dataset offer a variety of variables and correlations that are outside the purview of traditional analysis in order to comprehend the process of deriving meaning from the data. Since the data are highdimensional by nature, identifying the key variables influencing healthcare spending presents both a problem and an opportunity for sophisticated methodologies.
expenditures including patterns of the real gross domestic product (GDP) and health expenditures as a percentage of GDP for almost 200 countries in 2000 and 2022 . Principal component analysis was adopted to reduce dimensionality, thereby discerning the major expenditure-influencing factors and simplifying the complexity of the dataset. It aided in understanding variable importance, which was useful in further predictive modeling. Four ways for forecasting health expenditures were employed: AR (Autoregressive), MA (Moving Average), ARMA, LSTM (Long Short-Term Memory) networks. The AR and MA gave strong statistical assertions about past and recent trends, while ARMA was a hybrid method that combined autoregressive and moving average components to fit more complex time-dependent structures within the data. while the LSTM model learned the long-term dependency and non-linear relationships existing in the data. Two distinct LSTM models, Uni-variable (UV) and Multivariable (MV), were developed based on various indicators of healthcare expenditure. A comparison of these methods indicated the strength of each of the methods in context with performance, shedding light on how both traditional statistical methods and deep learning techniques may be successful in predictive analytics. Keywords—Principal Component Analysis, Dimensionality reduction, Healthcare Spending Patterns, Long Short-Term Memory, Autoregressive Model, Moving Average Model.
Due to the high dimensionality and complexity of the data presented in this study, they have been analysed using Principal Component Analysis (PCA) to reduce duplication without sacrificing the most significant factors affecting changes in CHE_GDP. PCA decreases the difficulty of data interpretation by naming key contributors to healthcare spending change trends while sustaining the fundamental structure of the data.
1. INTRODUCTION Health care spending acts as a key determinant for the capacity of the health system, the resilience of the economy, and the general state of public welfare in a nation. Funding streams in a health care system are fundamentally essential in shaping access to medical services, infrastructural development, and advancement in medical research. One of the most important financial indicators is Current CHE_GDP: Health Expenditure as a percentage of GDP; indicates the level of relative investment a country is making in health care. This study will focus on analysing and forecasting CHEGDP trends in about 200 countries starting from 2000 to 2022 to fathom a clear picture of the global health care financing framework.
© 2025, IRJET
|
Impact Factor value: 8.315
To perform the forecasting of future CHE_GDP values, four different predictive modeling techniques were applied Autoregressive (AR) Models Using past values of CHE_GDP to predict future trends based on time-dependent relationships. Moving Average (MA) Models are utilizing methods that capture short-lived deviations concerning healthcare expenditure from smooth past variations. Autoregressive and Moving Average (ARMA) Model with both autoregressive and moving average components, modeling current CHE_GDP based on past dependencies and short-run changes. Since then Long Short-Term Memory Networks (LSTMs) with Deep learning-based were
|
ISO 9001:2008 Certified Journal
|
Page 691