Peramalan Pertambahan Pasien Rawat Inap dengan Menggunakan Model Support Vector Regression (SVR)

Abstract

This research aims to see the prediction of the number of patients at the Medan Haji General Hospital in 2022-2023. A hospital is a health service institution that provides complete individual health services, providing inpatient, outpatient and emergency services. In implementing health services, hospitals must maintain medical records to support services and process patient information. This hospital serves all types of groups around North Sumatra. Prediction is the process of forecasting future demand which will include demand in terms of quantity, quality, time and location to meet demand for goods, services or the environment. This research uses the Support Vector Regression (SVR) method. Support Vector Regression (SVR) is a learning system that applies linear functions to a hypothetical feature space with high dimensions. The SVR algorithm concept can produce good forecasting values because SVR has the ability to solve overfitting problems. Overfitting is data behavior during the training phase that results in almost perfect prediction accuracy. Based on the results of data processing using the Support Vector Regression (SVR) method, it can be concluded that the application of the forecasting method in predicting the number of inpatient visits at RSU Haji Medan using the SVR method is carried out by determining predictions using three kernels, namely the RBF, linear and polynomial, then determine the best MSE and RMSE values to then be used as the best kernel. The results of forecasting inpatient visits using the SVR method show that the predicted results have decreased from the previous actual data which is not much different, but the predicted number of inpatients is almost the same every month and experiences insignificant decreases and increases.

Keywords: Prediction; Support Vector Regression (SVR).