Prathama, Aditya Heru
Unknown Affiliation

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

Hybrid Deep Learning Models for Free-Living Imbalanced Human Activity Recognition: Comparative Study Prathama, Aditya Heru; Joy Milliaan; Ghandy
Jurnal RESTI (Rekayasa Sistem dan Teknologi Informasi) Vol 10 No 2 (2026): April 2026
Publisher : Ikatan Ahli Informatika Indonesia (IAII)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29207/resti.v10i2.6936

Abstract

This study presents a comparative evaluation of three hybrid deep learning models for human activity recognition (HAR) in free-living and highly imbalanced conditions: 1DCNN-ResBLSTM-Attention (Model A), Attention-Mechanism-Based Deep Learning Feature Combination (Model B), and Time-Reversal-1DCNN-ResLSTM-Attention (Model C). Each architecture integrates convolutional layers for feature extraction, recurrent networks for temporal modeling, and attention mechanisms to enhance relevant representations. The HARTH v2.0 dataset, comprising 31 subjects and 15 activity classes under strong class imbalance, is used for evaluation. Results show that soft labeling consistently improves performance by better capturing transitional uncertainty in windowed sensor data. Model A achieves the highest accuracy (96.21%) and macro-averaged F1-score (88.17%), followed by Model C with comparable performance at lower computational cost, while Model B underperforms on minority classes due to limitations of spectrogram-based representations. Across all models, persistent confusion is observed among activities with similar motion patterns, such as walking, standing, and shuffling, indicating intrinsic ambiguity in sensor signals. This study provides a controlled and standardized comparison of hybrid architectures under realistic conditions, revealing both performance trade-offs and shared limitations. The findings highlight the importance of modeling uncertainty and temporal context for improving robustness, particularly transitional and underrepresented activities.
Rancang Bangun Sistem Pemantauan Kebisingan Berbasis IoT pada Jam Tenang di Asrama Mahasiswa Lalujan, Virginia; Prathama, Aditya Heru; Irawan, Jo'el Evander Nathanael; Christopher, Pixel Ariel; Silpinus, Davin Edbert
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 13 No 2: April 2026
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.132

Abstract

  Kesempatan tinggal di asrama selama masa studi membawa banyak manfaat bagi mahasiswa, seperti akses cepat ke fasilitas kampus, pengalaman hidup mandiri dalam komunitas, dan lingkungan yang tenteram untuk mendukung produktivitas akademik. Manfaat tersebut hanya dapat terwujud apabila ketenangan lingkungan asrama terjaga secara konsisten, khususnya pada periode jam tenang (quiet hour) yang diperuntukkan bagi belajar dan istirahat. Namun, pengelolaan kebisingan masih menjadi tantangan yang sering dikeluhkan oleh pamong asrama. Penelitian ini menyajikan rancang bangun sistem pemantauan kebisingan berbasis IoT untuk lingkungan asrama mahasiswa selama jam tenang. Sistem dirancang untuk pemantauan pada tingkat kamar dengan penempatan satu Noise Detection Kit (NDK) per kamar pada skenario target implementasi. NDK merupakan perangkat ringkas dan portabel sehingga memungkinkan implementasi simultan pada banyak titik. Pada tahap realisasi, sistem diimplementasikan secara terbatas pada sepuluh kamar terpilih sebagai sampel untuk memvalidasi kinerja dan fungsionalitas. Sistem mengintegrasikan sound sensor, NodeMCU, serta komunikasi MQTT dan Node-RED untuk visualisasi data melalui web dashboard dan mobile notifications via Telegram. Pada kondisi bising (>55 dB), indikator LED pada dashboard berubah menjadi merah disertai informasi kebisingan, log aktivitas, serta opsi aktivasi buzzer oleh pamong. Kebisingan kontinu lebih dari 5 detik memicu notifikasi otomatis ke ponsel pamong. Sensor dikalibrasi menggunakan Sound Level Meter (SLM) dengan MAE 1,02 dan RMSE 1,44, menunjukkan kinerja deteksi yang handal. Sistem ini memungkinkan pengelolaan kebisingan secara otomatis, objektif, real time, dan kontinu. Dibandingkan perangkat komersial sejenis atau SLM, sistem yang dikembangkan menawarkan solusi yang lebih ekonomis dengan skalabilitas yang lebih tinggi, serta peningkatan efisiensi seiring bertambahnya luas asrama yang dikelola.   Abstract Living in student dormitory during study period provides many benefits, including quick access to campus facilities, independent living experience within communities, and calm environment that supports academic productivity. These benefits can only be realized if dormitory environment remains consistently quiet, particularly during quiet hours intended for studying and resting. However, noise management remains a challenge frequently reported by dormitory supervisors. This study presents design and development of an IoT-based noise monitoring system for student dormitories during quiet hours. It is designed for room-level monitoring by deploying one Noise Detection Kit (NDK) per room in the target scenario. The NDK is compact and portable, enabling simultaneous deployment at multiple monitoring points. During implementation phase, the system was deployed on a limited scale in ten selected rooms as samples to validate system performance and functionality. The system integrates sound sensor, NodeMCU, MQTT, and Node-RED to enable data visualization through web-based dashboard and mobile notifications via Telegram. When noise levels exceed 55 dB, the dashboard LED indicator turns red and displays noise information, activity logs, and an option for supervisors to activate a buzzer. Continuous noise lasting more than 5 seconds triggers automatic notifications to the supervisor’s phone. The sensor was calibrated using a Sound Level Meter (SLM), achieving MAE of 1.02 and RMSE of 1.44, indicating reliable noise detection performance. The system enables automatic, objective, real-time, and continuous noise management. Compared to commercial devices or standalone SLMs, the proposed system is more economical and scalable, with increasing efficiency as dormitory area expands.