Comparison of Stock Price Forecasting with ARIMA and Backpropagation Neural Network (Case Study: Telkom Indonesia)

Katherine Liora Carissa; Betty Subartini; Sukono Sukono

doi:10.46336/ijqrm.v6i1.896

Comparison of Stock Price Forecasting with ARIMA and Backpropagation Neural Network (Case Study: Telkom Indonesia)

Katherine Liora Carissa, Betty Subartini, Sukono Sukono

Abstract

The growth of capital market investors in Indonesia is increasing every year. The most popular investment instrument is stocks. One of the stocks on the Indonesia Stock Exchange (IDX) is the Telkom Indonesia (TLKM). Through stock investment, investors can make a profit by utilizing stock prices in the market. However, stock price fluctuations are uncertain. Therefore, modeling is needed to be able to predict stock prices more accurately. The purpose of this study was to find an appropriate time series model and Neural Network model architecture, and to measure the accuracy of the two models in predicting future stock prices of TLKM. The study was conducted using the Autoregressive Integrated Moving Average (ARIMA) model and Backpropagation Neural Network (BPNN). For comparison, the Mean Absolute Percentage Error (MAPE) method was used. The data used in both models were the stock prices of Telkom Indonesia (TLKM) from September 1, 2023 to September 30, 2024. The result shows that the best ARIMA model, selected based on the least Akaike Information Criterion (AIC) value, is ARIMA(0,1,3) with a MAPE value of 1.20%. Meanwhile, the best BPNN model selected from the smallest testing Mean Squared Error (MSE) value, is BPNN(1,3,1) with a MAPE value of 1.17%. Among those two models, the BPNN model is more accurate because it has less MAPE value compared to the ARIMA one. The results of this research can be considered in forecasting TLKM stock price in the future.

Keywords

Stocks; ARIMA; Backpropagation; Neural Network; MAPE

Full Text:

PDF

References

Fausett, L. (1994). Fundamentals of Neural Networks: Architectures, Algorithms, and Applications. Hoboken: Prentice Hall, Inc.

Islam, M., Chen, G., & Jin, S. (2019). An Overview of Neural Network. American Journal of Neural Networks and Applications, 5(1), 7–11.

Kittichotsatsawat, Y., Boonprasope, A., Rauch, E., Tippayawong, N., & Tippayawong, K. Y. (2023). Forecasting arabica coffee yields by auto-regressive integrated moving average and machine learning approaches. AIMS Agriculture and Food, 8(4), 1052–1070.

Mustafidah, H., & Suwarsito, S. (2020). Performance of Levenberg-Marquardt Algorithm in Backpropagation Network Based on the Number of Neurons in Hidden Layers and Learning Rate. JUITA: Jurnal Informatika, 8(1), 29–35.

Pallant, J. (2011). SPSS SURVIVAL MANUAL. New York: Routledge.

Singh, A., Kushwaha, S., Alarfaj, M., & Singh, M. (2022). Comprehensive Overview of Backpropagation Algorithm for Digital Image Denoising. Electronics (Switzerland), 11(10), 1–15.

Sukono, Subartin, B., Ambarwati, Napitupulu, H., Saputra, J., & Hidayat, Y. (2019). Forecasting model of gross regional domestic product (GRDP) using backpropagation of Levenberg-Marguardt method. Industrial Engineering and Management Systems, 18(3), 530–540.

Taslim, D. G., & Murwantara, I. M. (2024). Comparative analysis of ARIMA and LSTM for predicting fluctuating time series data. Bulletin of Electrical Engineering and Informatics, 13(3), 1943–1951.

Wei, L., Yu, Z., Jin, Z., Xie, L., Huang, J., Cai, D., He, X., & Hua, X. S. (2019). Dual Graph for Traffic Forecasting. IEEE Access, 4, 1–10.

Wei, W. W. . (2006). Nonstationary Time Series Models. London: Pearson Education.

Works, R. (2021). An overview of activation functions used in Neuronal Networks: Application for plant disease diagnosis. International Journal of Advanced Trends in Computer Science and Engineering, 10(1), 139–145.

DOI: https://doi.org/10.46336/ijqrm.v6i1.896