Article Per Year (5 Year)
p-Index From 2021 - 2026
0.61
P-Index
This Author published in this journals
All Journal Jurnal Mantik Jurnal Locus Penelitian dan Pengabdian Online Learning in Educational Research
Sariyasa Sariyasa
Universitas Pendidikan Ganesha, Indonesia
Aerospace Engineering Automotive Engineering Computer Science & IT Decision Sciences, Operations Research & Management Economics, Econometrics & Finance Education Electrical & Electronics Engineering Languange, Linguistic, Communication & Media Mathematics

Published : 3 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 3 Documents
Search

 1 
Batch size and learning rate effect in covid-19 classification using CNN Ni Komang Rai Mirayanti; Sariyasa Sariyasa; I Gede Aris Gunadi
Jurnal Mantik Vol. 7 No. 3 (2023): November: Manajemen, Teknologi Informatika dan Komunikasi (Mantik)
Publisher : Institute of Computer Science (IOCS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35335/mantik.v7i3.4177

Abstract

This study evaluated the performance of a Convolutional Neural Network (CNN) model in classifying CT-Chest images of COVID-19 and non-COVID patients. The primary focus was to determine the influence of learning rate and batch size on the model's effectiveness. This research used 698 datas from Covid-19 and NonCovid-19 CT-Chest Image. Those dataset was obtained from medRxiv dan bioRxiv and has been approved by radiology expert in Tongji Hospital, China. In this research, COVID-16 dataset was classified by CNN with different batch sizes and learning rates for each iteration. Batch size used in this study were 1, 2, 4, 8, 16, 32, 64, and 128 with learning rates 0,00001; 0,0001; 0,001; 0,01; 0,1 and 1. This study found that this study showed that batch size and learning rate have a positive effect on CNN performance. The lower the learning rate, the lower the batch size will allow the network in the CNN model to perform better in classifying COVID-19 CT-Chest. In Addition, the best batch size for the classification is 64 with learning rate 0,01. These findings provided important insights into how parameters such as learning rate and batch size impacted the performance of the CNN model in classifying COVID-19 CT-Chest
Evaluasi Sistem Informasi Rapor SP Menggunakan Metode Pieces dan Metode Topsis (Studi Kasus SMKN 1 Kuta Selatan) Ni Kadek Erna Supriathi; I Gede Aris Gunadi; Sariyasa Sariyasa
Jurnal Locus Penelitian dan Pengabdian Vol. 5 No. 5 (2026): JURNAL LOCUS: Penelitian dan Pengabdian
Publisher : Riviera Publishing

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58344/locus.v5i5.5614

Abstract

Penilaian hasil belajar dalam pendidikan bertujuan untuk menyeimbangkan antara peningkatan capaian hasil belajar siswa dan proses perkembangan pembelajaran, dengan memanfaatkan perkembangan ilmu pengetahuan dan teknologi yang semakin maju guna meningkatkan kualitas pembelajaran. Rapor Satuan Pendidikan (Rapor SP), sebagai sistem informasi yang digunakan pada tingkat sekolah dasar, menengah pertama, menengah atas, kejuruan, dan pendidikan khusus, belum dimanfaatkan secara optimal. Sebagian guru masih memerlukan pendampingan dalam penggunaan sistem tersebut, dan pada periode tertentu seperti akhir semester, aplikasi sering mengalami kendala, seperti akses yang lambat serta proses login yang kurang efisien. Penelitian ini bertujuan untuk melakukan evaluasi secara komprehensif terhadap permasalahan implementasi Aplikasi E-Rapor SP di SMKN 1 Kuta Selatan dengan menggunakan metode PIECES untuk mengidentifikasi kelemahan sistem, serta menentukan dan memprioritaskan rekomendasi solusi perbaikan yang paling efektif dan efisien menggunakan metode TOPSIS. Teknik pengumpulan data meliputi kuesioner dengan skala Likert, wawancara, dan diskusi yang dianalisis menggunakan metode PIECES dan metode TOPSIS. Hasil penelitian menunjukkan bahwa prioritas utama rekomendasi perbaikan ditetapkan pada tiga solusi, yaitu penerapan sistem pencadangan data (backup) mendekati waktu nyata (near real-time) (A2), penguatan sosialisasi pengisian data email untuk proses reset password (A13), serta penerapan validasi format input secara otomatis (A10). Hasil analisis juga menunjukkan bahwa perbaikan pada infrastruktur sistem (stabilitas, keamanan, dan kualitas data) merupakan prasyarat yang lebih mendesak dan efektif (prioritas A2) dibandingkan dengan solusi berupa bimbingan teknis atau sosialisasi semata.
Bridging Verbal Geometry Problems and Formal Proofs: A GeoGebra-Supported Element–Relation Detection Framework I Putu Wisna Ariawan; I Made Ardana; I Made Sugiarta; Sariyasa Sariyasa; Raphita Yanisari Silalahi; Agus Adiarta; Dewa Gede Hendra Divayana
Online Learning In Educational Research (OLER) Vol. 6 No. 1 (2026): Online Learning in Educational Research
Publisher : CV FOUNDAE

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58524/oler.v6i1.1225

Abstract

Geometry proof problems presented without diagrams require students to transform verbal information into accurate visual representations before constructing formal deductive arguments. However, instructional approaches that explicitly support this transition remain limited. This study investigated the effectiveness of a GeoGebra-supported Element–Relation Detection (ERD) framework in strengthening undergraduate students’ plane geometry proof performance. A quasi-experimental design was conducted with first-year mathematics education students assigned to four instructional conditions combining the ERD framework, conventional Polya-based instruction, and the presence or absence of GeoGebra support. Students’ proof performance was evaluated through validated tasks measuring diagram construction, conjecture generation, proof organization, and argument validity. The findings indicate that learning environments supported by GeoGebra consistently produced stronger proof performance than conventional instruction without dynamic visualization. The ERD framework further assisted students in identifying essential geometric elements and relationships before constructing formal proofs, resulting in more coherent reasoning and fewer errors during the proof process. These findings suggest that the primary contribution of GeoGebra lies in facilitating accurate mathematical representations, while the ERD framework provides a structured pedagogical approach for bridging verbal geometric information and formal deductive reasoning. This study offers a practical instructional framework for improving geometry proof learning in technology-enhanced mathematics education.
Co-Authors Agus Adiarta Dewa Gede Hendra Divayana I Gede Aris Gunadi I Made Ardana I Made Sugiarta I Putu Wisna Ariawan Ni Kadek Erna Supriathi Ni Komang Rai Mirayanti Raphita Yanisari Silalahi