Claim Missing Document
Check
Articles

Found 7 Documents
Search

A Comparison of Retweet Prediction Approaches: The Superiority of Random Forest Learning Method Hendra Bunyamin; Tomas Tunys
TELKOMNIKA (Telecommunication Computing Electronics and Control) Vol 14, No 3: September 2016
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/telkomnika.v14i3.3150

Abstract

We consider the following retweet prediction task: given a tweet, predict whether it will be retweeted. In the past, a wide range of learning methods and features has been proposed for this task. We provide a systematic comparison of the performance of these learning methods and features in terms of prediction accuracy and feature importance. Specifically, from each previously published approach we take the best performing features and group these into two sets: user features and tweet features. In addition, we contrast five learning methods, both linear and non-linear. On top of that, we examine the added value of a previously proposed time-sensitive modeling approach. To the authors’ knowledge this is the first attempt to collect best performing features and contrast linear and non-linear learning methods. We perform our comparisons on a single dataset and find that user features such as the number of times a user is listed, number of followers, and average number of tweets published per day most strongly contribute to prediction accuracy across selected learning methods. We also find that a random forest-based learning, which has not been employed in previous studies, achieves the highest performance among the learning methods we consider. We also find that on top of properly tuned learning methods the benefits of time-sensitive modeling are very limited.
Masking preprocessing in transfer learning for damage building detection Hapnes Toba; Hendra Bunyamin; Juan Elisha Widyaya; Christian Wibisono; Lucky Surya Haryadi
IAES International Journal of Artificial Intelligence (IJ-AI) Vol 12, No 2: June 2023
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijai.v12.i2.pp552-559

Abstract

The sudden climate change occurring in different places in the world has made disasters more unpredictable than before. In addition, responses are often late due to manual processes that have to be performed by experts. Consequently, major advances in computer vision (CV) have prompted researchers to develop smart models to help these experts. We need a strong image representation model, but at the same time, we also need to prepare for a deep learning environment at a low cost. This research attempts to develop transfer learning models using low-cost masking pre-processing in the experimental building damage (xBD) dataset, a large-scale dataset for advancing building damage assessment. The dataset includes eight types of disasters located in fifteen different countries and spans thousands of square kilometers of satellite images. The models are based on U-Net, i.e., AlexNet, visual geometry group (VGG)-16, and ResNet-34. Our experiments show that ResNet-34 is the best with an F1 score of 71.93%, and an intersection over union (IoU) of 66.72%. The models are built on a resolution of 1,024 pixels and use only first-tier images compared to the state-of-the-art baseline. For future orientations, we believe that the approach we propose could be beneficial to improve the efficiency of deep learning training.
Model Prediksi Harga Penutupan di Bursa Efek Indonesia dengan Menggunakan Bidirectional LSTM dan HIVE-COTE Julian, Matthew; Bunyamin, Hendra
Jurnal Tekno Kompak Vol 19, No 1 (2025): FEBRUARI
Publisher : Universitas Teknokrat Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33365/jtk.v19i1.4599

Abstract

Memprediksi harga penutupan saham merupakan masalah yang menantang karena natur bursa saham yang volatil dan tidak linier. Adanya teknologi machine learning telah mendorong metode prediksi yang lebih akurat untuk data deret waktu, terutama dalam konteks prediksi harga saham. Prediksi yang akurat memberikan manfaat dalam bidang finansial, yaitu memudahkan investor dalam menganalisis pasar dan meminimalisir risiko. Model Bidirectional Long Short-Term Memory (Bi-LSTM) yang merupakan perkembangan dari Long Short-Term Memory (LSTM) dan Hierarchical Vote Collective of Transformation-based Ensembles 2.0 (HIVE-COTE 2.0) merupakan dua pendekatan yang digunakan dalam penelitian untuk memprediksi klasifikasi harga penutupan saham. Dataset yang digunakan adalah dataset harga saham yang berisikan informasi-informasi, seperti tanggal, harga pembukaan, harga tertinggi, harga terendah, harga penutupan, harga penutupan yang disesuaikan, dan volume penjualan. Kemudian, informasi-informasi ini dikenakan proses features engineering untuk memperoleh date/time features, lag features, rolling-window features, dan expanding-window features. Training model Bi-LSTM dan HIVE-COTE 2.0 dari fitur-fitur tersebut dan proses hyperparameter tuning dua model tersebut memberikan hasil evaluasi bahwa model HIVE-COTE 2.0 dengan performa F1-score 98.7% lebih baik dalam memprediksi klasifikasi harga saham dibandingkan dengan Bi-LSTM. Selanjutnya, analisis model untuk mencari fitur yang paling berpengaruh dengan dan tanpa feature engineering dilakukan  dan fitur yang paling berpengaruh adalah persentase perubahan harga pada akhir transaksi pada satu hari. Hasil penelitian ini merekomendasikan HIVE-COTE 2.0 sebagai model prediksi karena keakuratannya dan fitur persentase perubahan harga sebagai fitur yang paling mempengaruhi hasil prediksi.
Exploring Technology Integration in Education: Lecturers Perspective on Outcomes-Based Education Platforms Kasih, Julianti; Wasis, Galih Wicaksono; Bunyamin, Hendra
JOIV : International Journal on Informatics Visualization Vol 8, No 2 (2024)
Publisher : Society of Visual Informatics

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.62527/joiv.8.2.2691

Abstract

Informatics education is evolving rapidly through the adoption of Outcome-Based Education (OBE), necessitating a rigorous investigation into the effectiveness of the implementation. This study was conducted using the advanced Unified Theory of Acceptance and Use of Technology (UTAUT)-3 model to assess the potential of OBE systems in enhancing teaching and learning processes. The study integrated a comprehensive set of nine variables to measure the acceptance level of OBE systems among lecturers at Maranatha Christian University Bandung and Universitas Muhammadiyah Malang. UTAUT-3 provides a more explicit understanding by incorporating Hedonic Motivation (H.M.), Habit (H), and Personal Innovativeness (P.I.). The Model also integrated the core constructs of Performance Expectancy (P.E.), Effort Expectancy (E.E.), Social Influence (S.I.), Facilitating Conditions (F.C.), Behavioral Intention (B.I.), and Users Behavior (U.B.). The result showed that B.I. was a central determinant of U.B., suggesting users' preparedness to engage with OBE systems.Furthermore, the routine use of technology as Habit (H) was closely related to Behavioral Intension (B.I.), showing that familiarity with technology facilitated the intention to adopt OBE systems. The result showed that UTAUT-3's comprehensive framework was superior in evaluating educational technology adoption due to its ability to account for users' engagement as Hedonic Motivation (H.M.), dispositional tendencies toward Personal Innovativeness (P.I.), and the critical role of established habits. Consumers' actual experiences and technological proficiency significantly influence adoption rather than individual characteristics. Therefore, UTAUT-3 was a more effective tool for predicting and understanding the Acceptance of OBE systems, guiding educational institutions toward successfully integrating information systems in learning environments.
Utilizing Indonesian Universal Language Model Fine-tuning for Text Classification Bunyamin, Hendra
Journal of Information Technology and Computer Science Vol. 5 No. 3: Desember 2020
Publisher : Faculty of Computer Science (FILKOM) Brawijaya University

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (1985.283 KB) | DOI: 10.25126/jitecs.202053215

Abstract

Inductive transfer learning technique has made a huge impact on the computer vision field. Particularly, computer vision  applications including object detection, classification, and segmentation, are rarely trained from scratch; instead, they are fine-tuned from pretrained models, which are products of learning from huge datasets. In contrast to computer vision, state-of-the-art natural language processing models are still generally trained from the ground up. Accordingly, this research attempts to investigate an adoption of the transfer learning technique for natural language processing. Specifically, we utilize a transfer learning technique called Universal Language Model Fine-tuning (ULMFiT) for doing an Indonesian news text classification task. The dataset for constructing the language model is collected from several news providers from January to December 2017 whereas the dataset employed for text classification task comes from news articles provided by the Agency for the Assessment and Application of Technology (BPPT). To examine the impact of ULMFiT, we provide a baseline that is a vanilla neural network with two hidden layers. Although the performance of ULMFiT on validation set is lower than the one of our baseline, we find that the benefits of ULMFiT for the classification task significantly reduce the overfitting, that is the difference between train and validation accuracies from 4% to nearly zero.
PELATIHAN GURU DAN TANTANGAN BEBRAS 2024 UNTUK PENGENALAN COMPUTATIONAL THINKING DI BIRO BEBRAS MARANATHA Wijanto, Maresha Caroline; Toba, Hapnes; Ayub, Mewati; Karnalim, Oscar; Tan, Robby; Natasya, Rossevine Artha; Senjaya, Wenny Franciska; Adelia; Edi, Doro; Bunyamin, Hendra; Kasih, Julianti; Yulianti, Diana Trivena; Widjaja, Andreas; Johan, Meliana Christianti; Surjawan, Daniel Jahja; Zakaria, Teddy Marcus; Risal; Kandaga, Tjatur
Jurnal Abdimas Ilmiah Citra Bakti Vol. 6 No. 2 (2025)
Publisher : STKIP Citra Bakti

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.38048/jailcb.v6i2.5237

Abstract

Pemahaman siswa terhadap konsep Computational Thinking (CT) masih tergolong rendah, sementara pengenalan terhadap CT menjadi krusial di era digital saat ini. Tantangan Bebras menjadi sarana edukatif yang efektif untuk memperkenalkan CT melalui berbagai soal (Bebras task) yang bersifat aplikatif dan menantang. Kegiatan pengabdian masyarakat ini bertujuan untuk meningkatkan pemahaman dan keterlibatan siswa dalam CT melalui pembekalan guru dan pelaksanaan Tantangan Bebras 2024. Mitra kegiatan adalah guru dan siswa dari jenjang SD, SMP, dan SMA yang tergabung dalam Biro Bebras Maranatha. Metode yang digunakan meliputi lokakarya nasional, pelatihan guru, technical meeting, pelaksanaan Tantangan Bebras, dan evaluasi prestasi siswa. Hasil menunjukkan peningkatan partisipasi peserta sebanyak 4.429 siswa dari 136 sekolah, meningkat signifikan dibanding tahun sebelumnya. Sebanyak 165 siswa berhasil meraih peringkat 1–6, dengan sebagian besar berasal dari sekolah yang mengikuti Gerakan PANDAI. Evaluasi juga menunjukkan bahwa pembekalan guru efektif meningkatkan kesiapan dalam mengenalkan CT kepada siswa. Kegiatan ini menunjukkan bahwa kolaborasi antara pelatihan guru dan Tantangan Bebras dapat menjadi strategi efektif untuk memperluas pemahaman dan kemampuan siswa dalam CT.
Pembelajaran Computasional Thinking melalui Program Gerakan Pandai untuk Guru dan PKBM Ayub, Mewati; Wijanto, Maresha Caroline; Tan, Robby; Surjawan, Daniel Jahja; Toba, Hapnes; Christianti, Meliana; Edi, Doro; Bunyamin, Hendra; Adelia, Adelia; Risal, Risal
Aksiologiya: Jurnal Pengabdian Kepada Masyarakat Vol 7 No 3 (2023): Agustus
Publisher : Universitas Muhammadiyah Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30651/aks.v7i3.13430

Abstract

Program Gerakan Pandai yang digagas oleh Bebras Indonesia dengan dukungan Google bertujuan untuk membuat guru mulai menjadi guru penggerak dalam menyemaikan dan menumbuh-kembangkan kemampuan Computational Thinking (CT). Melalui gerakan PANDAI ini, diharapkan guru mengenal CT dan memperkenalkan CT kepada para siswa, sehingga siswa dapat mengembangkan kemampuan  berpikir komputasional yang bersifat kritis dan kreatif. Biro Bebras Maranatha menjalankan program Gerakan Pandai dalam dua batch yang dimulai pada bulan September 2020 sampai dengan Desember 2021. Pelatihan guru  batch1 diikuti oleh 148 guru, sedangkan batch2 diikuti 394 guru. Indikator guru yang berhasil menerapkan kemampuan CT adalah guru yang melaksanakan  paling sedikit 4 sesi microteaching dalam dua semester. Guru yang tuntas melakukan microteaching untuk batch1 ada 110 orang (74%), dan batch2 ada 184 guru (47%), dengan persentase rata-rata 60.5% untuk seluruh batch.Â