Claim Missing Document
Check
Articles

Found 2 Documents
Search
Journal : Indonesian Journal of Education And Computer Science

Fortifying Defenses: Exploring Innovative Security Strategies to Enhance Resilience and Safeguard IoT Networks from Emerging Threats Harahap, Ahmad Indra; Bahri, Syaiful; Figna, Harry Pratama
Indonesian Journal of Education And Computer Science Vol. 3 No. 1 (2025): INDOTECH - April 2025
Publisher : PT. INOVASI TEKNOLOGI KOMPUTER

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.60076/indotech.v3i1.1186

Abstract

The widespread adoption of the Internet of Things (IoT) has brought significant benefits to various sectors, but it has also introduced new challenges in terms of security and privacy. This article aims to elaborate on innovative security strategies that can be applied to enhance the resilience of IoT networks. Firstly, we conduct an in-depth review of the current threats faced by IoT networks, including hacking attacks, software exploits, and data misuse. Next, we introduce a series of innovative security strategies designed to address the identified risks. These strategies include the use of blockchain technology for distributed data security, the implementation of biometric-based authentication for user identification, and the utilization of machine learning for behavior-based attack detection. We also highlight the importance of security integration at every stage of the IoT device lifecycle, from design to implementation, as well as the need for cross-sector collaboration to build a holistic security ecosystem. By combining these strategies, it is expected that IoT networks can become more resilient and capable of addressing increasingly complex security threats in the future.
Evaluasi Kinerja GoogleNet Menggunakan Transfer Learning dan Fungsi Optimasi SGDM untuk Klasifikasi Citra Gulma Syechu, Weno; Syahputra, Rian; Harahap, Ahmad Indra
Indonesian Journal of Education And Computer Science Vol. 3 No. 1 (2025): INDOTECH - April 2025
Publisher : PT. INOVASI TEKNOLOGI KOMPUTER

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.60076/indotech.v3i1.1189

Abstract

Identifikasi gulma secara cepat dan tepat merupakan elemen penting dalam pertanian presisi. Penelitian ini memfokuskan pada evaluasi arsitektur Convolutional Neural Network (CNN) GoogleNet dalam klasifikasi citra gulma menggunakan pendekatan transfer learning. Dataset DeepWeeds yang berisi 17.509 gambar digunakan dan diklasifikasikan ke dalam sembilan kelas gulma. Proses pelatihan dilakukan dengan membekukan semua layer kecuali layer fully-connected terakhir, yang disesuaikan dengan jumlah kelas. Fungsi optimasi Stochastic Gradient Descent with Momentum (SGDM) digunakan dalam proses pelatihan. Penelitian ini mengevaluasi kinerja arsitektur Convolutional Neural Network (CNN) GoogleNet menggunakan pendekatan transfer learning untuk klasifikasi citra gulma pada dataset DeepWeeds yang terdiri dari sembilan kelas gulma berbeda. Fungsi optimasi Stochastic Gradient Descent with Momentum (SGDM) digunakan selama pelatihan model. Hasil eksperimen menunjukkan bahwa model mencapai akurasi pengujian sebesar 92,38% dengan waktu klasifikasi rata-rata hanya 0,0365 detik per gambar. Studi ini memberikan kontribusi signifikan sebagai acuan penerapan deep learning efisien dalam sistem pertanian presisi.