PERBANDINGAN METODE MACHINE LEARNING UNTUK PREDIKSI KUNJUNGAN PASIEN PADA RUMAH SAKIT SITTI KHADIDJAH GORONTALO

COMPARISON OF MACHINE LEARNING METHODS FOR PREDICTING PATIENT VISITS AT SITTI KHADIDJAH HOSPITAL GORONTALO

Abstract

The level of patient visits at RSIA Sitti Khadidjah Gorontalo has increased every year as seen from the data from 2020 to 2024. Therefore, researchers will predict the development of patient visits so that they can help RSIA Sitti Khadidjah in providing medical personnel, facilities, and medicines more effectively according to the estimated number of patients. This study aims to apply the Autoregressive Integrated Moving Average (ARIMA) and Neural Network methods in predicting patient visits and to find out the results of predicting patient visits at RSIA Sitti Khadidjah Gorontalo and the results of comparisons using the Autoregressive Integrated Moving Average (ARIMA) and Neural Network methods. The results of this study are the number of data records 60, then the smallest MSE value obtained from the ARIMA algorithm is found in the type of GENERAL Outpatient patient visits with an MSE value of 7.692 and the Neural Network algorithm with the smallest RMSE value is found in PBI Inpatient with an RMSE value of 0.069. Although the evaluation values ​​differ, ARIMA uses MSE while Neural Networks use RMSE. However, considering the smallest MSE and RMSE values, RMSE is superior due to its smaller value. Therefore, the Neural Network algorithm is superior to ARIMA.

Keywords: Prediction, Patient Visits, Autoregressive Integrated Moving Average, Neural Network.