Garuda - Garba Rujukan Digital

PERANCANGAN PENGUJIAN FUNGSIONAL DAN NON FUNGSIONAL APLIKASI SIAPPARA DI KABUPATEN HUMBANG HASUNDUTAN Riyanthi Angrainy Sianturi; Arnaldo Marulitua Sinaga; Yohanssen Pratama; Hotni Simatupang; Julio Panjaitan; Sandy Sihotang
J-Icon : Jurnal Komputer dan Informatika Vol 9 No 2 (2021): Oktober 2021
Publisher : Universitas Nusa Cendana

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35508/jicon.v9i2.4706

SIAPPARA is an application that used to manage market retribution by Dinas Koperasi, Perdagangan Dan Industri kabupaten Humbang Hasundutan. The SIAPPARA application consists of two platforms, namely a web application used by the admin and treasurer and a mobile application used by market officers. The application has been used by market officers in 12 people's markets in Humbang Hasundutan. To ensure that all functions in the application have been running according to the design and requirements, it is necessary to test both functional and non functional features. Testing is carried out with the aim of finding failures in applications that have been used. Before testing the application, the researcher makes a test design including the selection of methods and stages of testing. The method used in this research is literature study and experiment. Researchers used the literature study method in collecting theoretical references and information related to functional and non functional testing of applications. Researchers use the experimental method to choose the test method to be used and make a test design for the method that has been selected.The results of this study is a design of testing activities using Category Based Partition testing method, the Usability Testing test method with the Heuristic Evaluation technique, and the Performance Testing method.

PERANCANGAN APLIKASI ”SIAPPARA” UNTUK PELAPORAN SETORAN E-RETRIBUSI PASAR KABUPATEN HUMBANG HASUNDUTAN Verawaty Situmorang; Yohanssen Pratama; Riyanthi Angrainy Sianturi; Arnaldo Marulitua Sinaga
J-ICON : Jurnal Komputer dan Informatika Vol 9 No 2 (2021): Oktober 2021
Publisher : Universitas Nusa Cendana

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35508/jicon.v9i2.5256

Market retribution are local retribution that payed for the use of market facilities provided by the district government to individuals or entities. The process of collecting, recording, and reporting market retribution currently in Humbang Hasundutan Regency is still not effective and efficient, so it requires information technology support to enable it to improve the quality of services that are transparent and accountable in the traditional markets of Humbang Hasundutan Regency. The applications that will be developed are mobile and web-based applications that allow the collection and recording process to be more effective and efficient as well as more transparent reporting. Through the “SIAPPARA” Application, it is hoped that the process of collecting retribution in several markets in Humbang Hasundutan Regency will be easier and more transparent.

PENGEMBANGAN APLIKASI WEB PARIWISATA DANAU TOBA BERBASIS KOMUNITAS (VISIT TOBA) Yohanssen Pratama; Riyanthi A Sianturi; Helmuth S Tampubolon; Kristopel Lumbantoruan; Hotni M Simatupang; Indah T Tampubolon; Yohana C Manullang
J-ICON : Jurnal Komputer dan Informatika Vol 10 No 2 (2022): Oktober 2022
Publisher : Universitas Nusa Cendana

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35508/jicon.v10i2.7867

The problem faced by several tourist destinations in the Lake Toba Region is that branding has not been maximized using digital media. In this study, we have the goal of developing the VisitToba Community Based Tourism (CBT) application to develop the tourism potential of the Lake Toba area through digitalization with the concept of community empowerment, improving the quality of tourism and the economy of the tourism community in the Lake Toba area. The Visit Toba application is expected to be a forum for the development of local small and medium-sized businesses as tourism supporters, where tourism actors can take advantage of the features in this application to be able to collaborate with each other to make a tour package. Increased community participation in efforts to develop and strengthen tourism in the Lake Toba area is the expected result with this application, so that a partnership is established between tourism entrepreneurs in the Lake Toba area with the tour packages provided. Applications built using the Waterfall model and have passed the testing phase so that the results in the form of a website can already be implemented in the Lake Toba area.

FEW-SHOT LEARNING FOR AML CELL CLASSIFICATION USING PROTOTYPICAL NETWORKS Dirgayussa, I Gde Eka; Herman, Kevin Elfancyus; Nugroho, Doni Bowo; Sekar Asri Tresnaningtyas; Meita Mahardianti; Nurul Maulidiyah; Rafli Filano; Rudi Setiawan; Muhammad Artha Jabatsudewa Maras; Yohanssen Pratama
METHODIKA: Jurnal Teknik Informatika dan Sistem Informasi Vol. 11 No. 2 (2025): Volume 11 Nomor 2 Tahun 2025
Publisher : Universitas Methodist Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.46880/mtk.v11i2.4650

Accurate blood cell classification is crucial for diagnosing Acute Myeloid Leukemia (AML) but limited medical data poses challenges for traditional machine learning models. This study presents a Few-Shot Learning (FSL) framework utilizing a Prototypical Network architecture with a ResNet-34 backbone to classify AML blood cell types from microscopic images. In this study, we utilize datasets consisting of 15 morphologically distinct cell classes. A 15-way, 5-shot, 5-query episodic setup was adopted to simulate data-scarce conditions. Evaluation via 5-fold cross-validation yielded strong performance, with an average accuracy of 97.76%, precision of 98.78%, recall of 96.55%, and F1-score of 97.76%. FSL training times were consistent (4.22–4.26 minutes per fold), and t-SNE along with confusion matrices confirmed the model’s ability to distinguish similar cell types. To validate the approach, its performance was compared with a conventional supervised CNN using the same ResNet-34 backbone. The FSL model outperformed the CNN across all metrics such as accuracy (98.32% vs. 77.25%), precision (98.55% vs. 76.87%), recall (98.31% vs. 78.66%), and F1-score (98.33% vs. 75.26%), while also requiring far less training time (~4.24 min/fold vs. ~420 min total). These results highlight the promise of FSL based methods for accurate, efficient, and scalable hematologic diagnostics in data limited settings.

FEW-SHOT LEARNING FOR AML CELL CLASSIFICATION USING PROTOTYPICAL I Gde Eka Dirgayussa; Kevin Elfancyus Herman; Doni Bowo Nugroho; Sekar Asri Tresnaningtyas; Meita Mahardianti; Nurul Maulidiyah; Rafli Filano; Rudi Setiawan; Muhammad Artha Jabatsudewa Maras; Yohanssen Pratama
METHODIKA: Jurnal Teknik Informatika dan Sistem Informasi Vol. 11 No. 2 (2025): Volume 11 Nomor 2 Tahun 2025
Publisher : Universitas Methodist Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar

Accurate blood cell classification is crucial for diagnosing Acute Myeloid Leukemia (AML) but limited medical data poses challenges for traditional machine learning models. This study presents a Few-Shot Learning (FSL) framework utilizing a Prototypical Network architecture with a ResNet-34 backbone to classify AML blood cell types from microscopic images. In this study, we utilize datasets consisting of 15 morphologically distinct cell classes. A 15-way, 5-shot, 5-query episodic setup was adopted to simulate data-scarce conditions. Evaluation via 5-fold cross-validation yielded strong performance, with an average accuracy of 97.76%, precision of 98.78%, recall of 96.55%, and F1-score of 97.76%. FSL training times were consistent (4.22–4.26 minutes per fold), and t-SNE along with confusion matrices confirmed the model’s ability to distinguish similar cell types. To validate the approach, its performance was compared with a conventional supervised CNN using the same ResNet-34 backbone. The FSL model outperformed the CNN across all metrics such as accuracy (98.32% vs. 77.25%), precision (98.55% vs. 76.87%), recall (98.31% vs. 78.66%), and F1-score (98.33% vs. 75.26%), while also requiring far less training time (~4.24 min/fold vs. ~420 min total). These results highlight the promise of FSL based methods for accurate, efficient, and scalable hematologic diagnostics in data limited settings.

Title

Found 5 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 5 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 5 Documents
Search