---

HYBRID AUTOREGRESSIVE INTEGRATED MOVING AVERAGE AND NEURAL NETWORK AUTOREGRESSIVE METHODS (CASE STUDY OF INDONESIA SHARIA STOCK DATA)

Rizkah Novirah Latunrung, Georgina Maria Tinungki*, Nirwan

---

Download PDF

---

Abstract

The Indonesian Sharia Stock Index (ISSI) was first officially introduced on May 12, 2011. ISSI data is a fluctuating time series data which is something that is uncertain and difficult to predict. therefore forecasting is one of the important things to do. Good forecasting indicators produce the right forecasting value, a forecasting method is needed that is in accordance with the characteristics or patterns of the data. In general, there are two types of time series data patterns, namely linear and nonlinear, linear patterns using the ARIMA method and nonlinear using NNAR. The results of the ARIMA-NNAR hybrid are the best model with an RMSE of 1.099 and a MAPE value of 4.375%. The forecasting results of the Indonesian Sharia Stock Index on April 21, 2025 to April 30, 2025 are 206.0341; 207.1324; 208.3564; 210.4519; 212.1144; 213.4942; 215.0161; 216.646, respectively.

Keywords: Forecasting, Time Series, ARIMA, NNAR, Indonesian Sharia Stock Index (ISSI)