The master’s thesis addresses the issue of rising costs in the healthcare sector of the EU. Member states have been grappling with escalating healthcare expenditures for decades, which is a consequence of an aging population, rapid advancements in
medical technology, the development of financial regulations in the healthcare sector and increasing healthcare prices, the later is known as medical inflation.
I have presented three models for predicting time series data and used them to forecast medical inflation in EU countries. Firstly, I used the well-established autoregressive integrated moving average model (ARIMA model), which was then improved by adding a seasonal element to create the seasonal autoregressive integrated
moving average model (SARIMA model). Secondly, I used the modern recurrent neural network model (RNN model). I compared and evaluated the accuracy of medical inflation predictions for different forecasting horizons. I discovered that all models adequately capture the complexity of the examined time series dynamics, but the ARIMA and SARIMA models are more suitable for short-term predictions of medical inflation, while the RNN model provides more accurate forecasts for periods longer than six months.
Accurate predictions of medical inflation would enable government officials, healthcare providers, and insurance companies to proactively address the challenges of rapid development of the healthcare sector, thus significantly contributing to ensuring
financially accessible and high-quality healthcare for EU residents.
|
