Claim Missing Document
Check
Articles

Found 33 Documents
Search

Optimization of Convolutional Neural Network for Classification of Hydroponic Vegetable Cultivation Using Machine Learning Lubis, Arif Ridho; Prayudani, Santi; Putra, Purwa Hasan; Lase, Yuyun Yusnida
Journal of Applied Engineering and Technological Science (JAETS) Vol. 7 No. 1 (2025): Journal of Applied Engineering and Technological Science (JAETS)
Publisher : Yayasan Riset dan Pengembangan Intelektual (YRPI)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37385/jaets.v7i1.7231

Abstract

In an effort to apply applied product innovation and support the improvement of hydroponic vegetable cultivation, it is based on several things. Among them are changes in the texture of the year, stems and vegetable quality. At this time the problems faced by hydroponic vegetable pickers, especially banyumas village youth organizations who have UMKM hydroponic vegetable cultivation. This situation will have an impact on problems and losses that result in a lack of yield and quality of harvested vegetables if not resolved quickly. The results of this study resulted in optimal accuracy performance in the classification of hydroponic vegetables with CNN, this study also successfully classified normal vegetables with vegetables affected by disease. This research produces accuracy in the first test 73% and the second test 92%.
PENGEMBANGAN APLIKASI BASIS INFORMASI PENELITIAN DAN PENGABDIAN KEPADA MASYARAKAT PADA P3M POLITEKNIK NEGERI MEDAN Putra, Purwa Hasan; Lase, Yuyun Yusnida; Asmara, Wira Bayu
JOURNAL OF SCIENCE AND SOCIAL RESEARCH Vol 9, No 1 (2026): February 2026
Publisher : Smart Education

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54314/jssr.v9i1.5832

Abstract

Abstract: This study aims to analyze the utilization of the Simlibtamas system in managing research proposal submissions at the Medan State Polytechnic. Based on observations on the system dashboard, there are 1,299 registered users, 21 of whom are active, who have submitted proposals, and a total of 800 proposals that have been processed. This system is able to display information in a structured manner through a submission table containing the name of the proposer, the proposal title, the original scheme, document files, assessments, and the amount of proposed and recommended funding. In addition, a recapitulation graph of the amount of funding shows the existence of funding submissions within a certain time period, reflecting the dynamics of research activity. The results show that the Simlibtamas system has functioned well in supporting the administration and evaluation process of research proposals. However, the involvement of proposer users still needs to be improved for more optimal system utilization. In conclusion, Simlibtamas is able to facilitate monitoring, transparency, and management of research, and can be further developed to increase participation and effectiveness of research fund distribution. Keywords: Information Systems, Applications, Simlibtimnas, Research, Community Service Abstrak: Penelitian ini bertujuan untuk menganalisis pemanfaatan sistem Simlibtamas dalam pengelolaan pengajuan proposal penelitian di Politeknik Negeri Medan. Berdasarkan hasil pengamatan pada dashboard sistem, tercatat sebanyak 1299 pengguna terdaftar dengan 21 pengguna aktif yang mengajukan proposal, serta total 800 proposal yang telah diproses. Sistem ini mampu menampilkan informasi secara terstruktur melalui tabel pengajuan yang memuat nama pengusul, judul proposal, skema pendanaan, file dokumen, penilaian, serta jumlah dana yang diusulkan dan direkomendasikan. Selain itu, grafik rekapitulasi jumlah dana menunjukkan adanya fluktuasi pengajuan dana dalam rentang waktu tertentu, yang mencerminkan dinamika aktivitas penelitian. Hasil penelitian menunjukkan bahwa sistem Simlibtamas telah berfungsi dengan baik dalam mendukung proses administrasi dan evaluasi proposal penelitian. Namun, keterlibatan pengguna pengusul masih perlu ditingkatkan agar pemanfaatan sistem menjadi lebih optimal. Kesimpulannya, Simlibtamas mampu memudahkan monitoring, transparansi, dan pengelolaan penelitian, serta dapat dikembangkan lebih lanjut untuk meningkatkan partisipasi dan efektivitas distribusi dana penelitian. Kata kunci: Sistem Informasi, Aplikasi, Simlibtimnas, Penelitian, Pengabdian
Pemanfaatan Canny Edge Detection untuk Pembacaan OMR Survey 7 Kebiasaan Anak Indonesia Hebat Friendly Friendly; Harizahayu Harizahayu; Purwa Hasan Putra
INSOLOGI: Jurnal Sains dan Teknologi Vol. 5 No. 1 (2026): Februari 2026
Publisher : Yayasan Literasi Sains Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55123/insologi.v5i1.6705

Abstract

The 7 Kebiasaan Anak Indonesia Hebat program is one of the priority program of the Ministry of Primary and Secondary Education of the Republic of Indonesia. It aims to develop students with strong academic ability, good behavior, and strong character. The program is implemented in all primary and secondary schools, and thus large amounts of data must be summarized for monthly reports. Urban schools often use Google Forms or digital applications to record students’ activities, while schools in smaller towns or rural areas still rely on paper forms. Limited access to smartphones and internet connections among parents makes online data collection difficult. Consequently, teachers must manually summarize data using spreadsheet applications, increasing their workload—especially when managing many students. This study proposes the use of Canny Edge Detection to automate data processing from OMR (Optical Mark Recognition) sheets. By scanning or photographing filled OMR sheets, the system can accurately read and convert students’ responses into digital data. This method allows teachers to digitize the reporting process and reduce manual work. Using this method, the accuracy of reading the OMR sheets can reach 81% while the need of informing the parents to fill the form correctly since some tested data shown recall data reached 68%.