Perancangan Sistem Prediksi Pola Permintaan Layanan Internet FTTH Menggunakan Model LSTM Berbasis Web
DOI:
https://doi.org/10.36312/tjm6wa36Keywords:
FTTH, Long Short-Term Memory, prediksi permintaan, sistem berbasis web, time seriesAbstract
Penelitian ini bertujuan merancang sistem berbasis web untuk memprediksi pola permintaan layanan internet Fiber to the Home (FTTH) dengan memanfaatkan model Long Short-Term Memory (LSTM) sebagai komponen analitik. Permasalahan yang dihadapi penyedia layanan FTTH adalah fluktuasi permintaan antarwaktu dan antarwilayah yang berdampak pada perencanaan stok perangkat, penjadwalan pemasangan, dan pengembangan infrastruktur. Penelitian menggunakan pendekatan Cross Industry Standard Process for Data Mining (CRISP-DM) dengan data historis permintaan layanan FTTH periode 2022 sampai 2025 yang diagregasi menjadi data runtun waktu bulanan. Sistem dikembangkan menggunakan Python, Flask, dan SQLite, sedangkan model LSTM dikonfigurasi dengan dua layer, 50 unit neuron, optimizer Adam, dan 50 epoch. Hasil penelitian menunjukkan bahwa sistem mampu mengintegrasikan pengelolaan data pelanggan, proses prediksi, visualisasi tren, dan rekomendasi kebutuhan stok dalam satu platform. Pada evaluasi awal, nilai loss pelatihan menurun dari 0.8541 menjadi 0.0064, sedangkan evaluasi prediksi menghasilkan Mean Absolute Error sebesar 32.43 dan Root Mean Squared Error sebesar 32.68. Temuan ini menunjukkan bahwa model telah mampu mengikuti pola umum permintaan pada data yang digunakan, tetapi kinerja operasionalnya masih perlu divalidasi lebih lanjut melalui pembandingan dengan model baseline, skema evaluasi time-series yang lebih ketat, dan konteks volume permintaan per wilayah. Kontribusi utama penelitian ini terletak pada integrasi hasil prediksi ke dalam sistem operasional FTTH berbasis web, bukan pada pengusulan arsitektur LSTM baru.
This study aimed to design a web-based system to predict Fiber to the Home (FTTH) internet service demand patterns using a Long Short-Term Memory (LSTM) model as the analytical component. The main challenge faced by FTTH providers is the fluctuation of demand across time periods and service areas, which affects equipment stock planning, installation scheduling, and infrastructure expansion. The study adopted the Cross Industry Standard Process for Data Mining (CRISP-DM) and used historical FTTH service demand data from 2022 to 2025, aggregated into monthly time-series data. The system was developed using Python, Flask, and SQLite, while the LSTM model was configured with two layers, 50 neurons, the Adam optimizer, and 50 epochs. The results show that the system integrates customer data management, prediction processes, trend visualization, and stock requirement recommendations within a single platform. In the initial evaluation, the training loss decreased from 0.8541 to 0.0064, while the prediction evaluation yielded a Mean Absolute Error of 32.43 and a Root Mean Squared Error of 32.68. These findings indicate that the model captured the general demand pattern in the available data, although its operational validity still requires further verification through baseline comparisons, stricter time-series evaluation schemes, and demand-scale context. The main contribution of this study lies in integrating forecasting outputs into an FTTH operational web system rather than proposing a new LSTM architecture.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Shalshabila Shafa Putry, Yeka Hendriyani, Yulia Fatmi, Syafrijon Syafrijon

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with Journal of Authentic Research agree to the following terms:
- For all articles published in Journal of Authentic Research, copyright is retained by the authors. Authors give permission to the publisher to announce the work with conditions. When the manuscript is accepted for publication, the authors agrees to implement a non-exclusive transfer of publishing rights to the journals.
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution-ShareAlike 4.0 International License that allows others to share the work with an acknowledgment 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 acknowledgment 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 (See The Effect of Open Access).

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.