Claim Missing Document
Check
Articles

Found 20 Documents
Search

Adaptive Inertia Weight Particle Swarm Optimization for Augmentation Selection in Coral Reef Classification with Convolutional Neural Networks Prabowo, Dwi Puji; Rohman, Muhammad Syaifur; Megantara, Rama Aria; Pergiwati, Dewi; Saraswati, Galuh Wilujeng; Pramunendar, Ricardus Anggi; Shidik, Guruh Fajar; Andono, Pulung Nurtantio
JOIV : International Journal on Informatics Visualization Vol 9, No 1 (2025)
Publisher : Society of Visual Informatics

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.62527/joiv.9.1.2726

Abstract

Indonesia possesses the world's largest aquatic resources, with 17,504 islands and 6.49 million square kilometers of sea. Located in the coral triangle, Indonesia is home to diverse marine life, including vital coral reefs. However, these reefs face threats from climate change, pollution, and human activities, endangering biodiversity and coastal communities. Therefore, monitoring and preservation are crucial. This study evaluates various augmentation methods for classifying underwater coral reef images using Convolutional Neural Networks (CNNs). Effective augmentation methods are essential due to the unique characteristics of these images. The methodology includes testing different augmentation methods, epoch parameters, and CNN parameters on a coral reef image dataset. Five optimization algorithms (AIWPSO, GA, GWO, PSO, and FOX) are compared. The highest accuracy, 95.64%, is achieved at the 10th epoch. AIWPSO and GA show the highest average accuracies, 93.44%, and 93.50%, respectively, with no significant performance differences among the algorithms. Statistical analysis using the Wilcoxon test indicates a significant difference between training and validation accuracy (p-value = 0.0020). These findings underscore the importance of selecting augmentation methods that align with the characteristics of each optimization algorithm to enhance classification performance. The results provide valuable insights into improving the quality and diversity of input data for classification algorithms in underwater image analysis. They highlight the necessity of matching augmentation methods to specific optimization algorithms to boost accuracy and effectiveness significantly. Future research should explore additional augmentation methods and optimization algorithms further to enhance the robustness and accuracy of underwater image classification.
A Study of Governance and Public Participation in Indonesian Megaprojects: A Comparative Analysis with International Practices Rohman, Muhammad Syaifur
Jurnal Planologi Vol 22, No 1 (2025): April
Publisher : Universitas Islam Sultan Agung Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30659/jpsa.v22i1.43458

Abstract

This research analyzes governance and public participation in infrastructure megaprojects in Indonesia, such as MRT Jakarta and Trans-Java Toll Road, by comparing national and international practices. Megaprojects frequently face challenges in the form of cost overruns, schedule delays, social conflicts, and significant environmental impacts. This study aims to evaluate the effectiveness of implemented governance and public participation models, and compare them with international best practices. The study uses a literature approach and international case analysis to explore governance models, such as Public-Private Partnership (PPP) and more inclusive public participation approaches. The results show that transparent, accountable governance that involves communities from the outset can enhance public acceptance and project sustainability. Comparative analysis with international case studies provides insights into adapting best practices to the Indonesian context. Strategic recommendations include financing innovations, utilization of digital technology, and a holistic approach that considers social, economic, and environmental impacts.Keywords: Governance, Public Participation, Megaprojects, Infrastructure, Sustainability
Prediksi Perubahan Tutupan Lahan di Kabupaten Bogor Tahun 2026 Menggunakan Random Forest dengan Citra Satelit Sentinel-2 Terklasfikasi Rohman, Muhammad Syaifur; Afrinaldi, Afrinaldi; Syauqani, Ahmad; Safira, Maya
Tunas Agraria Vol. 8 No. 2 (2025): Tunas Agraria
Publisher : Diploma IV Pertanahan Sekolah Tinggi Pertanahan Nasional

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31292/jta.v8i2.413

Abstract

Bogor Regency has experienced significant land cover changes due to urbanization, population growth, and infrastructure expansion. This study predicts land cover changes in 2026 using a random forest model based on classified Sentinel-2 satellite imagery. The model was trained with data from 2017, 2020, and 2023 and evaluated using 2-fold time-series cross-validation, with an accuracy of 87.96%, Kappa 0.8131, and F1-Score 0.8752. The prediction results show an increase in built-up area from 748.02 km² (2017) to 953.89 km² (2023) and is estimated to reach 976.84 km² in 2026—especially in Pamijahan and Jonggol. On the other hand, agricultural areas decreased from 652.53 km² to 541.11 km² and are predicted to decrease again to 530.33 km², threatening local food security. Tree cover areas also decreased from 1,509.12 km² (2017) to 1,385.34 km² (2023) but are expected to increase to 1,413.42 km² in 2026 due to the reforestation program. These findings emphasize the importance of sustainable land planning to balance development with environmental conservation for the sustainability of the ecosystem and the welfare of the Bogor community.   Kabupaten Bogor mengalami perubahan tutupan lahan yang signifikan akibat urbanisasi, pertumbuhan penduduk, dan ekspansi infrastruktur. Penelitian ini memprediksi perubahan tutupan lahan tahun 2026 menggunakan model Random Forest berbasis citra satelit Sentinel-2 yang telah diklasifikasi. Model dilatih dengan data tahun 2017, 2020, dan 2023, serta dievaluasi menggunakan 2-fold time-series cross-validation, dengan akurasi 87,96%, Kappa 0,8131, dan F1-Score 0,8752. Hasil prediksi menunjukkan peningkatan area terbangun dari 748,02 km² (2017) menjadi 953,89 km² (2023), dan diperkirakan mencapai 976,84 km² pada 2026—terutama di Pamijahan dan Jonggol. Sebaliknya, area pertanian menurun dari 652,53 km² menjadi 541,11 km², dan diprediksi turun lagi menjadi 530,33 km², mengancam ketahanan pangan lokal. Area tutupan pohon juga menurun dari 1.509,12 km² (2017) ke 1.385,34 km² (2023), namun diperkirakan meningkat menjadi 1.413,42 km² pada 2026 karena program reboisasi. Temuan ini menegaskan pentingnya perencanaan lahan berkelanjutan untuk menyeimbangkan pembangunan dengan pelestarian lingkungan, demi keberlanjutan ekosistem dan kesejahteraan masyarakat Bogor.
Location Based Service for improving Chabot Disaster Management Evacuator Palu Rohman, Muhammad Syaifur
Jurnal Transformatika Vol. 18 No. 1 (2020): July 2020
Publisher : Jurusan Teknologi Informasi Universitas Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26623/transformatika.v18i1.1890

Abstract

The devastating earthquake that struck Palu on the island of Sulawesi last September ripped through the Earth's crust at a rare high speed, scientists have found. When the disaster is over, many natural disaster victims need immediate help. The call center provided is usually busy with services and complaints from victims of natural disasters. The greater the impact of natural disasters, the more information services that must be carried out. By using CEPAT chatbot for disaster evacuation in Palu, information about evacuation place can be given to victim by access it. Then when the victim shares their location, CEPAT will give the nearest evacuation place information using LBS improvement Chabot.
Penerapan Metode Naïve Bayes Classifier Untuk Klasifikasi Sentimen Pada Judul Berita Astuti, Yani Parti; Wibowo, Alrico Rizki; Kartikadarma, Etika; Subhiyakto, Egia Rosi; Sri Winarsih, Nurul Anisa; Rohman, Muhammad Syaifur
LogicLink Vol. 1 No. 1, June 2024
Publisher : Universitas Islam Negeri K.H. Abdurrahman Wahid Pekalongan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.28918/logiclink.v1i1.7684

Abstract

News has a major role as a source of information to convey reports on opinions, events, and the latest findings in various aspects of life. News headlines, as an important component, can be a determinant of news content. The sentiment contained in news headlines can be classified using sentiment analysts, as is the case in the online media platform Kompas.TV. News headlines are retrieved using an automated program that utilises the HTML body with the help of NodeJs as the technology for program creation. This research is focused on the application of Naïve Bayes Classifier method to classify sentiment on Kompas.TV news headlines in Semarang City. The results showed an accuracy rate of 91.04%, with a ratio of training data and test data of 90:10. The conclusion of this study is that the Naïve Bayes Classifier method is effective in identifying news headlines with negative sentiment on Kompas.TV, with a precision of 89% and recall of 94%. This finding makes a positive contribution to the understanding of sentiment analysis on news headlines in online media, especially in the context of Kompas.TV news in Semarang City.
Prediksi Kerusakan Bangunan Pasca Gempa Bumi Menggunakan Metode Deep Neural Network Fakhrurrozi, Fakhrurrozi; Ratmana, Danny Oka; Winarsih, Nurul Anisa Sri; Saraswati, Galuh Wilujeng; Rohman, Muhammad Syaifur; Saputra, Filmada Ocky; Pramunendar, Ricardus Anggi; Shidik, Guruh Fajar
Jurnal Teknologi Sistem Informasi dan Aplikasi Vol. 7 No. 1 (2024): Jurnal Teknologi Sistem Informasi dan Aplikasi
Publisher : Program Studi Teknik Informatika Universitas Pamulang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32493/jtsi.v7i1.37181

Abstract

Addressing the challenge of predicting earthquake-induced building damage, this study proposes the innovative use of Deep Neural Networks (DNN) as a solution. Focusing on optimizing predictive models, the research evaluates the effectiveness of various optimizers - ADAM, SGD, RMSprop, and Adagrad - coupled with adjustments in the learning rate to determine the most efficient configuration. The experiment was conducted to compare the performance of each optimizer in predicting post-earthquake building damage, a critical issue in disaster mitigation. The results demonstrate that ADAM significantly outperforms other optimizers, achieving the highest accuracy of up to 90.50% at a learning rate of 0.001, with RMSprop as its closest competitor. While SGD and Adagrad yielded lower accuracies, SGD showed improvement with higher learning rates. The variance analysis confirmed that the choice of optimizer significantly impacts model performance, with the p-value indicating strong statistical significance for optimizers (1.23E-09), whereas the learning rate had no significant impact (p-value 0.56098964). These findings underline the importance of selecting the appropriate optimizer to enhance the accuracy of DNN models for building damage prediction, a crucial aspect in emergency response planning and earthquake disaster mitigation efforts. This research contributes significantly to the development of more accurate predictive models, which are essential in minimizing the risks of earthquake disasters.
Implementasi Algoritma Floyd Warshall Pada Aplikasi Dewan Masjid Indonesia (Dmi) Kota Semarang Untuk Menentukan Masjid Terdekat Rohman, Muhammad Syaifur; Saraswati, Galuh Wilujeng; Winarsih, Nurul Anisa Sri
Jurnal Informatika: Jurnal Pengembangan IT Vol 8, No 3 (2023)
Publisher : Politeknik Harapan Bersama

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30591/jpit.v8i3.4895

Abstract

Location Based Service (LBS) is a service on smartphones that functions as a navigation device based on the user's position to determine the location where the user is. LBS utilizes GPS capabilities in finding geolocation information and sometimes using Google maps to display a complete map of the location. But the results of previous research studies Google Map does not give shortest and accessible routes. Furthermore, to improve work of LBS, Floyd Warshall algorithm is used because the algorithm has the principle of optimality in calculating the total of all routes optimally. According to data recorded by the Ministry of Religion of the Republic of Indonesia there have been 1,304 Mosques in the City of Semarang, but with this much data it should be easier to find places of worship for Muslims. Most mosques that are visited are mosques on the highway because it is more visible even though there are many other mosques that can be accessed. By using the White Box and Black Box tests, finding shortest path to find places of worship in the city of Semarang can be given accurately. The result was the Floyd Warshall algorithm could provide shortest path route and it was more accessible better than Google Map navigation.
Pelatihan Implementasi Artificial Intelligence Menggunakan Teachable Machine berbasis Project-Based Learning bagi Siswa SMA/SMK Wibowo, Dibyo Adi; Hidajat, Moch. Sjamsul; Pramunendar, Ricardus Anggi; Rohman, Muhammad Syaifur; Ratmana, Danny Oka; Megantara, Rama Aria
ABDIMASKU : JURNAL PENGABDIAN MASYARAKAT Vol 9, No 1 (2026): JANUARI 2026
Publisher : LPPM UNIVERSITAS DIAN NUSWANTORO

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.62411/ja.v9i1.3226

Abstract

Artificial Intelligence (AI) merupakan teknologi yang berkembang pesat dan penting untuk dikenalkan sejak jenjang pendidikan menengah. Kegiatan pengabdian kepada masyarakat ini bertujuan untuk meningkatkan pemahaman siswa SMA/SMK di Kota dan Kabupaten Kediri terhadap konsep dasar Artificial Intelligence dan machine learning melalui pelatihan implementasi AI menggunakan Teachable Machine berbasis Project-Based Learning (PjBL). Metode pelaksanaan kegiatan mengombinasikan pendekatan PjBL dan experiential learning, di mana peserta dilibatkan secara aktif dalam pengembangan proyek AI sederhana berbasis gambar, suara, dan pose tubuh. Evaluasi pembelajaran dilakukan menggunakan pre-test dan post-test untuk mengukur peningkatan pemahaman peserta. Hasil kegiatan menunjukkan adanya peningkatan yang signifikan pada seluruh kategori materi, termasuk konsep dasar AI, computational thinking, machine learning, penggunaan Teachable Machine, serta implementasi dan evaluasi model AI. Temuan ini menunjukkan bahwa penggunaan Teachable Machine yang dipadukan dengan pendekatan PjBL efektif dalam meningkatkan literasi Artificial Intelligence siswa SMA/SMK serta membantu peserta memahami konsep AI secara lebih konkret dan aplikatif.
PENDAMPINGAN DIGITALISASI KEUANGAN DAN PEMASARAN DIGITAL UNTUK MENINGKATKAN KAPASITAS USAHA PADA UMKM RUANG AUDIO Al Afthoni, Insyirohul Qulub; Al Abrori, Muhammad Nabil; Prasetya, Bagus Hari; Fais, Muhammad; Rohman, Muhammad Syaifur
JMM (Jurnal Masyarakat Mandiri) Vol 10, No 1 (2026): Februari
Publisher : Universitas Muhammadiyah Mataram

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31764/jmm.v10i1.37232

Abstract

Abstrak: UMKM Ruang Audio merupakan usaha pembuatan miniatur sound system yang memiliki potensi pengembangan usaha, namun masih menghadapi permasalahan pencatatan keuangan manual dan pemasaran digital yang belum optimal. Kegiatan pengabdian ini bertujuan meningkatkan kapasitas pengelolaan usaha melalui digitalisasi keuangan dan pemasaran. Metode pelaksanaan mencakup observasi awal dan wawancara untuk mengidentifikasi permasalahan mitra, perancangan sistem pencatatan keuangan digital, pembuatan akun pemasaran digital, serta pelatihan teknis berbasis praktik langsung yang dilanjutkan dengan monitoring dan evaluasi. Pendampingan melibatkan 1 pemilik usaha dan 10 karyawan sebagai peserta. Implementasi kegiatan meliputi penerapan sistem pencatatan keuangan berbasis Google Sheets, pembuatan akun Facebook Marketplace “Ruangau Dio”, serta pendaftaran lokasi produksi Ruang Audio pada Google Maps. Evaluasi dilakukan melalui observasi selama kegiatan dan evaluasi lanjutan satu minggu pascapendampingan. Hasil kegiatan menunjukkan peningkatan hardskill mitra sebesar 90%, peningkatan soft skill pemasaran digital sebesar 80%, serta peningkatan aksesibilitas lokasi usaha sebesar 100% melalui Google Maps.Abstract: Ruang Audio, a small and medium-sized enterprise (SME), manufactures miniature sound systems and has the potential for business development, but still faces challenges with manual financial recording and suboptimal digital marketing. This community service activity aims to improve business management capacity through financial and marketing digitalization. The implementation method includes initial observation and interviews to identify partner issues, design a digital financial recording system, create a digital marketing account, and provide hands-on technical training followed by monitoring and evaluation. The mentoring program involved one business owner and 10 employees. Implementation included the implementation of a Google Sheets-based financial recording system, the creation of a "Ruangau Dio" Facebook Marketplace account, and the registration of Ruang Audio's production location on Google Maps. Evaluation was conducted through observations during the activity and a follow-up evaluation one week after the mentoring. The results of the activity showed a 90% increase in partners' hard skills, an 80% increase in digital marketing soft skills, and a 100% increase in business location accessibility via Google Maps.
Optimasi Model Transfer Learning menggunakan Sequential Hyperparameter Tuning pada Klasifikasi Huruf Bisindo Fawaid, Moh Adzka; Pramunendar, Ricardus Anggi; Winarsih, Nurul Anisa Sri; Rohman, Muhammad Syaifur; Ratmana, Danny Oka
Simetris: Jurnal Teknik Mesin, Elektro dan Ilmu Komputer Vol. 17 No. 1 (2026): JURNAL SIMETRIS VOLUME 17 NO 1 TAHUN 2026
Publisher : Fakultas Teknik Universitas Muria Kudus

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Klasifikasi huruf alfabet Bahasa Isyarat Indonesia (Bisindo) menjadi fokus penting dalam bidang computer vision. Variasi posisi tangan, pencahayaan, dan kemiripan antar huruf menyebabkan kesulitan dalam mengenali pola visual Bisindo. Penelitian ini menerapkan Convolutional Neural Network (CNN) berbasis Transfer Learning (TL) untuk menganalisis pengaruh hyperparameter terhadap kinerja model menggunakan metode Coordinate Descent Search. Empat arsitektur diuji, yaitu EfficientNetB0, EfficientNetB1, ResNet18, dan ResNet50, pada dataset Bisindo yang berisi 6.760 citra tangan huruf A–Z. Optimasi dilakukan secara sekuensial pada lima parameter: batch size, learning rate, dropout rate, optimizer, dan epoch. Hasil menunjukkan learning rate optimal antara 0.0001–0.0003 dengan batch size kecil (16) efektif untuk model ringan, sedangkan ResNet50 optimal pada batch size 64. Model ResNet50 menghasilkan akurasi 93,17%, precision 93,26%, recall 93,15%, dan F1-score 93,13%. Hal ini menunjukkan bahwa TL efektif untuk klasifikasi Bisindo, dan arsitektur lebih dalam seperti ResNet50 memberikan performa terbaik.