Hidayah, Rahmi Nur and Siregar, Riki Ruli Affandi (2026) PENERAPAN MODEL LONG SHORT-TERM MEMORY (LSTM) UNTUK PERAMALAN TREN LABA BERSIH BERDASARKAN AKTIVITAS TRANSAKSI BANK MALUKU. Diploma thesis, Institut Teknologi PLN.
Cover.pdf
Download (157kB)
Lembar Pengesahan.pdf
Restricted to Registered users only
Download (549kB)
Abstrak.pdf
Restricted to Registered users only
Download (214kB)
BAB I.pdf
Download (255kB)
BAB II.pdf
Download (655kB)
Bab V Kesimpulan.pdf
Restricted to Registered users only
Download (197kB)
Daftar Pustaka.pdf
Download (169kB)
BAB IV.pdf
Restricted to Registered users only
Download (557kB)
BAB III.pdf
Restricted to Registered users only
Download (1MB)
202231056_Rahmi Nur Hidayah_Revisi_Skripsi.pdf
Restricted to Registered users only
Download (5MB)
Abstract
Peramalan tren laba bersih merupakan aspek penting di sektor perbankan karena
mencerminkan kinerja keuangan dan efektivitas operasional bank. Namun, fluktuasi aktivitas
transaksi, dinamika ekonomi makro, serta karakteristik data deret waktu yang nonlinier
menyebabkan metode statistik konvensional belum optimal dalam merepresentasikan
hubungan antar periode secara menyeluruh, sehingga berpotensi memengaruhi ketepatan
dalam penentuan kebijakan dan perencanaan keuangan. Berdasarkan kondisi tersebut,
diperlukan model yang mampu mengakomodasi kompleksitas dinamika temporal data laba
bersih secara lebih akurat. Untuk mengatasi permasalahan ini, penelitian ini bertujuan untuk
menerapkan model Long Short-Term Memory (LSTM) dalam meramalkan tren laba bersih
Bank Maluku periode 2022–2024. Dataset yang digunakan dibagi menjadi 80% data
pelatihan dan 20% data pengujian. Hasil penelitian menunjukkan bahwa model LSTM
mampu menghasilkan prediksi yang sangat mendekati data aktual. Performa terbaik dicapai
pada pengujian ke-3 dengan nilai Root Mean Squared Error (RMSE) sebesar 24.290.518,18,
Mean Absolute Error (MAE) sebesar 18.043.777,07, Mean Absolute Percentage Error
(MAPE) sebesar 9,70%, dan koefisien determinasi (R2) sebesar 0,84304. Nilai MAPE di
bawah 10% menunjukkan tingkat akurasi yang tinggi, sementara nilai R2 yang signifikan
membuktikan kemampuan model dalam menjelaskan variasi data aktual secara mendalam
Net profit trend forecasting is a crucial aspect of the banking sector as it reflects financial
performance and operational effectiveness. However, fluctuations in transaction activities,
macroeconomic dynamics, and the non-linear characteristics of time-series data render
conventional statistical methods suboptimal in comprehensively representing inter-period
relationships. This limitation potentially affects the accuracy of policy determination and
financial planning. Consequently, a model capable of accurately accommodating the
complex temporal dynamics of net profit data is required. To address this issue, this study
aims to apply the Long Short-Term Memory (LSTM) model to forecast the net profit trends of
Bank Maluku for the 2022–2024 period. The dataset was divided into 80% training data and
20% testing data. The results demonstrate that the LSTM model is capable of generating
predictions that closely align with the actual data. The best performance was achieved in the
third test, yielding a Root Mean Squared Error (RMSE) of 24,290,518.18, a Mean Absolute
Error (MAE) of 18,043,777.07, a Mean Absolute Percentage Error (MAPE) of 9.70%, and a
coefficient of determination (R2) of 0.84304. A MAPE value below 10% indicates a high level
of accuracy, while the significant R2 value proves the model's ability to profoundly explain
the variation in actual data.
| Item Type: | Thesis (Diploma) |
|---|---|
| Uncontrolled Keywords: | Long Short-Term Memory, Peramalan, Laba Bersih, Deret Waktu, Perbankan Long Short-Term Memory, Forecasting, Net Profit, Time Series, Banking |
| Subjects: | Bidang Keilmuan > Data Mining Bidang Keilmuan > Deep learning Skripsi Bidang Keilmuan > Teknik Informatika |
| Divisions: | Fakultas Telematika Energi > S1 Teknik Informatika |
| Depositing User: | Mrs Hidayah Rahmi Nur |
| Date Deposited: | 04 Mar 2026 07:12 |
| Last Modified: | 04 Mar 2026 07:12 |
| URI: | https://repository.itpln.ac.id/id/eprint/5667 |
