HERDIAWAN, REZA DWI (2025) PERBANDINGAN ALGORITMA ARTIFICIAL NEURAL NETWORK DAN LONG SHORT TERM MEMORY DALAM PREDIKSI NILAI TUKAR RUPIAH KE YUAN. S1 thesis, Universitas Mercu Buana Jakarta.
|
Text (HAL COVER)
01 Cover.pdf Download (373kB) | Preview |
|
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (BAB I)
02 Bab 1.pdf Restricted to Registered users only Download (135kB) |
|
|
Text (BAB II)
03 Bab 2.pdf Restricted to Registered users only Download (301kB) |
|
|
Text (BAB III)
04 Bab 3.pdf Restricted to Registered users only Download (199kB) |
|
|
Text (BAB IV)
05 Bab 4.pdf Restricted to Registered users only Download (286kB) |
|
|
Text (BAB V)
06 Bab 5.pdf Restricted to Registered users only Download (30kB) |
|
|
Text (DAFTAR PUSTAKA)
07 Daftar Pustaka.pdf Restricted to Registered users only Download (149kB) |
|
|
Text (LAMPIRAN)
08 Lampiran.pdf Restricted to Registered users only Download (986kB) |
Abstract
This study compares the performance of Artificial Neural Network (ANN) and Long Short-Term Memory (LSTM) algorithms in predicting the exchange rate of Rupiah against Yuan, whose fluctuations are influenced by global economic factors. The main challenge is to capture non-linear and temporal patterns in complex historical data. Using a quantitative approach, this study analyzes 4,926 data from 2006 to 2024, which have been processed through normalization and feature selection. The ANN model was trained with Multi-Layer Perceptron architecture and ReLU activation function, while the LSTM used layers with 50 units and dropout to prevent overfitting. Performance evaluation based on Mean Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Squared Error (RMSE) shows that LSTM has higher accuracy, with MAE of 0.0011 and RMSE of 0.0019, compared to ANN which has MAE of 0.0013 and RMSE of 0.0023. The LSTM also obtained an R² of 0.9999, superior to the ANN with an R² of 0.9998. This study confirms the effectiveness of LSTM for long-term time series data, making a significant contribution to economic policy makers and financial market participants through the application of artificial intelligence-based models to support data-based decision making. Kata kunci: Exchange rate prediction, Artificial Neural Network (ANN), Long Short-Term Memory (LSTM), Rupiah, Yuan Penelitian ini membandingkan kinerja algoritma Artificial Neural Network (ANN) dan Long Short-Term Memory (LSTM) dalam memprediksi nilai tukar Rupiah terhadap Yuan, yang fluktuasinya dipengaruhi oleh faktor ekonomi global. Tantangan utama adalah menangkap pola non-linear dan temporal dalam data historis yang kompleks. Menggunakan pendekatan kuantitatif, penelitian ini menganalisis 4.926 data dari tahun 2006 hingga 2024, yang telah diproses melalui normalisasi dan seleksi fitur. Model ANN dilatih dengan arsitektur Multi-Layer Perceptron dan fungsi aktivasi ReLU, sedangkan LSTM menggunakan lapisan dengan 50 unit dan dropout untuk mencegah overfitting. Evaluasi performa berdasarkan Mean Absolute Error (MAE), Mean Squared Error (MSE), dan Root Mean Squared Error (RMSE) menunjukkan bahwa LSTM memiliki akurasi lebih tinggi, dengan MAE sebesar 0,0011 dan RMSE sebesar 0,0019, dibandingkan ANN yang memiliki MAE sebesar 0,0013 dan RMSE sebesar 0,0023. LSTM juga memperoleh R² sebesar 0,9999, lebih unggul dari ANN dengan R² sebesar 0,9998. Penelitian ini menegaskan efektivitas LSTM untuk data deret waktu jangka panjang, memberikan kontribusi signifikan bagi pengambil kebijakan ekonomi dan pelaku pasar keuangan melalui penerapan model berbasis kecerdasan buatan untuk mendukung pengambilan keputusan berbasis data. Kata kunci: Prediksi nilai tukar, Artificial Neural Network (ANN), Long ShortTerm Memory (LSTM), Rupiah, Yuan
Actions (login required)
 |
View Item |
