Robusta London Coffee Price Forecasting Analysis Using Recurrent Neural Network – Long Short Term Memory (RNN – LSTM)
DOI:
https://doi.org/10.26623/transformatika.v20i2.5482Keywords:
Coffee price forecasting, LSTM, Units, Dropout, RMSE, MAP.Abstract
Coffee price forecasting has a significant role in preventing price fluctuations at a time. Therefore, a method is needed that can be used to forecast the price of coffee. This study discusses the analysis of coffee price forecasting using the Recurrent Neural Network – Long Short-Term Memory (RNN – LSTM) method. This study will be determined the best LSTM model that aims to get the results of forecasting the price of London robusta coffee with the smallest Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE) values. Using the LSTM model with units of 128 and dropouts of 0.1, forecasting the price of London robusta coffee has an RMSE value of 1,303 and MAPE of 3.53%. Therefore, the LSTM model can indicate the cost of London robusta coffee with an accuracy rate of 96.47%.
References
Utami, T.W. and Darsyah, M.Y. Peramalan Data Saham dengan Model Winter’s. Jurnal Statistika. 2015; 3(2). https://doi.org/10.26714/jsunimus.3.2.2015.%25p
Rahardjo, P. Kopi. Jakarta : Penebar Swadaya Grup. 2012.
Gao, T., Chai, Y., dan Liu, Y. Applying Long Short Term Memory Neural Networks for Predicting Stock Closing Price. Proceeding IEEE International Conference on Software Engineering and Service Science (ICSESS). 2017; 575-578.
Susanti, R., & Adji, A. R. Analisis Peramalan IHSG dengan Time Series Modeling ARIMA. Jurnal Manajemen Kewirausahaan. 2020; 17: 97-106.
Wiranda, L., and Sadikin, M. Penerapan Long Short Term Memory Pada Data Time Series untuk Memprediksi Penjualan Produk PT. Metiska Farma. Jurnal Nasional Pendidikan Teknik Informatika. 2019; 8(3): 184-196. https://dx.doi.org/10.23887/janapati.v8i3.19139
Metode Peramalan untuk Identifikasi Potensi Permintaan Konsumen. Informatics Journal. 2019; 4(3): 121-129.
Han, J., Kamber, M., dan Pei, J. Data Mining: Concepts and Techniques. 3rd Edition. San Francisco: Elsevier. 2012.
Kotu, V. dan Deshpande, B. Predictive Analytics and Data Mining: Concepts and Practice with RapidMiner. Waltham: Elsevier. 2015
Pustejovsky, J. dan Stubbs, A. Natural Language Annotation for Machine Learning. California: O'Reilly Media. 2012.
Prathama, A.Y., Aminullah, A., dan Saputra, A. Pendekatan ANN (Artificial Neural Network) Untuk Penentuan Prosentase Bobot Pekerjaan Dan Estimasi Nilai Pekerjaan Struktur Pada Rumah Sakit Pratama. Teknosains. 2017; 7(1): 14-25.
Tian, C., Ma, J., Zhang, C., dan Zhan, P. A Deep Neural Network Model for Short Term
Load Forecast Based on Long Short-Term Memory Network and Convolutional Neural Network.
Energies. 2018; 11.
Qiu, J., Wang, B., dan Zhou, C. Forecasting Stock Prices with Long-Short Term Memory Neural Network Based on Attention Mechanism. Advanced Design and Intelligent Computing. 2020; 15(1).
Lewis, N. D. Neural Network for Time Series Forecasting with R. US: CrateSpace
Independent Publishing Platform. 2017.
Vercellis, C. Business Intelligence: Data Mining and Optimization for Decision Making. United
Kingdom: John Wiley & Son. 2009.
Budiman, H. Analisis Dan Perbandingan Akurasi Model Prediksi Rentet Waktu Support
Vector Machines Dengan Support Vector Machines Particle Swarm Optimization Untuk Arus Lalu
Lintas Jangka Pendek. Systemic. 2016; 2(1): 19-24.
Downloads
Additional Files
Published
Issue
Section
License
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work.
Transformatika is licensed under a Creative Commons Attribution 4.0 International License.
