Fatma Agus Setyaningsih, Sri Andayani, Retno Subekti
Investment in Islamic stock assets is currently increasingly in demand by the public. Similar to investing in stocks, this particular investment involves a substantial amount of risk because of the potential for rapid price fluctuations. Consequently, a forecasting tool is required to assist investors in thinking twice before acquiring Islamic stock. The data used is daily data from 2019 to 2022 with a total data of around 1200. The machine learning approaches we selected are variants of the Recurrent Neural Network model, namely Elman recurrent neural network (ERNN), long short-term memory (LSTM), and gated recurrent unit (GRU). The results of GRU models using mean absolute error (MAE) value is 0.0203. The root mean square error (RMSE)is 0.0325 in the GRU model with the best combination of hyperparameters. The model can make prediction values with small error values based on this combination of a proportion of 70% training data and 30% test data. This study recommends that the model is better to use more variations in hidden neurons, layers, activation functions, training algorithms, and parameters to get a better model architecture. © Little Lion Scientific.
Department of Mathematics, Faculty of Mathematics and Natural Science, Yogyakarta State University, Indonesia
