Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal IJAIT (International Journal of Applied Information Technology) Jurnal Impresi Indonesia
Artwodini Muqtadiroh, Feby
Unknown Affiliation
Religion Computer Science & IT Economics, Econometrics & Finance Education Law, Crime, Criminology & Criminal Justice Social Sciences

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

Found 2 Documents
Search

 1 
Ensemble Measurement Model of E-Learning Implementation Readiness for Higher Education Institution Artwodini Muqtadiroh, Feby; Ma'ady, Mochamad Nizar Palefi; Rizal Rahman, Faisal
IJAIT (International Journal of Applied Information Technology) Vol 02 No 02 (November 2018)
Publisher : School of Applied Science, Telkom University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25124/ijait.v2i02.1206

Abstract

The low level of student readiness in implementing e-learning can achieve not optimal benefits or even generate losses. In fact, its failures are able to impact on swelling of the institutional funds. Therefore, it is so important to measure the level of student readiness for avoiding the impact of e-learning implementation failures. In this study, we employed an ensemble model for measuring e-learning readiness level by using the model of Akaslan & Law and Aydin & Tasci. The data were obtained from questionnaires based on the factor of technology, people, content, institutions, acceptance for e-learning, and training for e-learning. With a questionnaire consisting of questions and five Likert scales, the survey was conducted in several departments who have implemented e-learning. We assessed the results by Akaslan & Law to measure the level of student readiness and Aydin & Law for determining its readiness By assessing the level of difficulty of e-learning implementation, we show a different way of assessment that the model of Akaslan & Law can also be used for the same variables related to Aydin & Tasci model in the measurement. We work in hand together between the model of Akaslan & Law and Aydin & Tasci for the sake of knowing the level of e-learning implementation readiness in such a different way. We observed e-learning of Institut Teknologi Sepuluh Nopember (ITS) or so-called ShareITS and found that ITS students are ready to implement e-learning.
Mengkuantifikasi Trade-off Biaya-Kualitas dalam Autoscaling Kubernetes Berbasis Reinforcement Learning Rohmat, Rohmat; Hery Purnomo, Mauridhi; Artwodini Muqtadiroh, Feby
Jurnal Impresi Indonesia Vol. 5 No. 2 (2026): Jurnal Impresi Indonesia
Publisher : Riviera Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58344/jii.v5i2.7527

Abstract

Penelitian autoscaling Kubernetes berbasis reinforcement learning saat ini kurang memiliki evaluasi statistik yang ketat, dengan sebagian besar hanya mengandalkan eksperimen skenario tunggal yang tidak dapat membedakan keunggulan algoritmik yang sebenarnya dari variasi acak. Kesenjangan metodologis ini melemahkan validitas dan reprodusibilitas peningkatan kinerja yang dilaporkan dalam literatur manajemen sumber daya cloud. Studi ini mengembangkan kerangka evaluasi berbasis simulasi yang kuat secara statistik untuk membandingkan secara ketat algoritma autoscaling reinforcement learning terhadap Horizontal Pod Autoscaler (HPA) Kubernetes standar, membangun metodologi benchmarking yang dapat direproduksi dengan pengujian signifikansi statistik yang tepat dan kuantifikasi ukuran efek. Sebuah simulator kejadian diskrit Python yang mengemulasi komponen control-plane Kubernetes (Metrics Server, Controller Manager, Scheduler) dengan dinamika siklus hidup pod yang realistis telah dikembangkan. Autoscaler Hybrid DQN-PPO dan HPA dievaluasi menggunakan desain eksperimen berpasangan di 30 skenario lalu lintas sintetis independen selama 24 jam. Analisis statistik menggunakan uji normalitas Shapiro-Wilk, koreksi Holm-Bonferroni untuk perbandingan berganda, ukuran efek Cohen’s d, dan interval kepercayaan bootstrap. Hasil mengungkapkan trade-off fundamental antara biaya dan kualitas: Hybrid DQN-PPO mencapai kualitas layanan superior dengan 60,58% lebih sedikit pelanggaran SLA, 19,61% lebih cepat latensi P95, dan 4,83% lebih cepat waktu respons rata-rata (semua p < 0, 001). Namun, peningkatan kualitas ini memerlukan premi biaya 8,92% ($6,87 per skenario) dibandingkan dengan HPA, yang mempertahankan utilisasi CPU 7,96% lebih tinggi melalui efisiensi sumber daya yang agresif (p < 0, 001). Perbedaan kinerja berasal dari strategi kontrol yang sangat berbeda: HPA menggunakan kontrol reaktif (menunggu pelanggaran sebelum scaling), mengoptimalkan biaya; Hybrid menggunakan kontrol prediktif (mencegah pelanggaran secara proaktif).
Co-Authors Mauridhi Hery Purnomo Mochamad Nizar Palefi Ma’ady Rizal Rahman, Faisal Rohmat rohmat