Admin Website UNIKOM

Tag: campus AI ready

  • Prediction Models of Acceptance Student Scholarships in University for Students K-12 Using Decision Tree Learning

    Authors: Agus Nursikuwagus, Tono Hartono, Agus Setiana, Muhammad Agil Alfariski, Refaldi Satria Gumelar, Muhammad Rizdky Maulady

    DOI: 10.1109/INCITEST64888.2024.11121501

    Abstract

    Receiving scholarships for high school students is one of the key processes in obtaining scholarships at a university; therefore, the process of monitoring and evaluating scholarship recipients at the high school level is very necessary. We haven’t used machine learning for prediction yet. A fair and suitable process is crucial to support the justification candidate. This research aims to provide prediction learning models based on the proposed dataset. We leveraged classification techniques like decision trees base, naive bayes learning, and support vector learning. As a result of the task, we received a variety of accuracy levels, ranging from tree learning to support vector machine. Each model presents accuracy values of 1.00,1.00, 1.00,0.81, and 0.99, in that order. SVM outperforms the other models, particularly in reducing false predictions. The dataset consists of 547 instances, with 70% of them in the training dataset and 30% in the testing dataset. Future research can leverage deep learning methods or large language machines. Providing the tuning parameter is crucial for enhancing classification accuracy. © 2024 IEEE.

    Author keywords

    classification learning; confusion matrix; decision tree; models; prediction

    This article can be accessed at https://www.scopus.com/pages/publications/105015870242

  • The Utilization of Data Visualization in Klinik Keluarga within the Cianjur District

    Authors: Alif Finanditha, Dian Dharmayanti, Juliadit Syahputra

    DOI: 10.1109/INCITEST64888.2024.11121481

    Abstract

    The objective of this research is to develop effective and efficient visualizations to assist the Promkes & Marketing Coordinator, and Warehouse Coordinator in acquiring the necessary information in a more quick and precise manner to improve the decision-making process. Based on the results of interviews conducted with the Promkes & Marketing Coordinator, the Warehouse Coordinator said that the current form of information presentation is a graph. However, this form of presentation has not helped the Promkes & Marketing Coordinator and Warehouse Coordinator perform their duties and misinterpret the information. With so much data that needs to be processed, this results in the Promkes & Marketing Coordinator and Warehouse Coordinator taking longer to get information, which hinders the tasks of the Promkes & Marketing Coordinator and Warehouse Coordinator. In this research, the method used adapts from 7 steps of data visualization adapted to the research to build a good visualization according to the needs. The knowledge extraction process is performed using descriptive statistics and data mining techniques. The implementation of the visualization design is done through the creation of a prototype, which is then tested through the usability testing method. The test results indicate that the data visualization prototype enables health promotion & marketing and warehouse coordinators to grasp the necessary information more swiftly. © 2024 IEEE.

    Author keywords

    data mining; data visualization; usability testing

    This article can be accessed at https://www.scopus.com/pages/publications/105015855737 

  • Exhaust Gas Detection System in Cabin Car Based on Internet of Things

    Authors: Dedena Hirawan, Audy Revi Nugraha

    DOI: 10.1109/INCITEST64888.2024.11121491

    Abstract

    Driving a car is a daily transportation activity carried out by the Indonesian people. Currently, there is no feature on the car to check and provide warnings about the air quality in the cabin. The purpose of this study is to build an exhaust gas detection system in the car cabin by analyzing the air using the MQ-135 and MQ-7 sensors. This study uses a prototyping model for software and hardware development. The prototyping model is quite effective at every stage, from data collection stage to system testing and implementation stage. This study focuses on developing a system to detect exhaust gases inside car cabins, as there is currently no feature in cars to monitor cabin air quality. The system uses MQ-135 and MQ-7 sensors to analyze air quality and detect harmful gases such as Carbon Monoxide (CO), Carbon Dioxide (CO2), Ammonia (NH3), and Nitrogen Oxides (NOx), which result from fossil fuel combustion. If gas levels exceed safe limits, the system provides a warning and recommends opening the windows to improve air circulation. The goal is to help maintain healthy air quality in car cabins, preventing negative health impacts on passengers, and allowing car owners to monitor vehicle conditions for safe daily use. © 2024 IEEE.

    Author keywords

    Car Cabin; Detection System; Exhaust Gas; Internet of Things

    This article can be accessed at https://www.scopus.com/pages/publications/105015892910

  • Utilization of Database Replication for Performance Improvement and Backup of Higher Education Academic Information System Database

    Author: Andri Heryandi

    DOI: 10.1109/INCITEST64888.2024.11121494

    Abstract

    Currently, a higher education institution has an Academic Information System to support the implementation of educational processes. Over time, as the information system is used, the amount of stored data also increases. This leads to slower database services, which subsequently affects the performance of the information system. In terms of security, a database must also be prepared to face potential attacks or disruptions that could be damaging. Therefore, the database used in an information system must support steps to minimize the impact of such attacks or disruptions, for example, through backup measures. However, implementing backup procedures in a database can impact the overall performance of the information system, such as causing downtime during the backup process. This research creates a database server architecture that can enhance database performance (availability, speed, and security), thereby improving the performance of the academic information system while also supporting the backup process without compromising the performance of the system information using database replication technology. © 2024 IEEE.

    Author keywords

    academic information system; backup; database; replication; security

    This article can be accessed at https://www.scopus.com/pages/publications/105015890567

  • Tempeh Chips Production Optimization System using the Simplex Method

    Authors: Gentisya Tri Mardiani, Rina Susanto

    DOI: 10.1109/INCITEST64888.2024.11121428 

    Abstract

    In producing tempeh chips, the company has limitations such as the availability of raw materials and daily production capacity. The simplex method is used to calculate the optimal amount of production in one day and find out how many products are produced. The aim of this research is to assist Production Managers in calculating the amount of production time carried out and knowing the optimal number of products produced from each production. Based on the results of this analysis has been implemented then the application of the Simplex Method can be implemented to produce 1x production of salty tempeh chips and 2x production of sweet tempeh chips with a total production of 3500 kilos per day so that the company can maximize the use of raw materials, machines and production capacity and optimize the number of products produced. The system to be built has three users, namely Production Manager, Production Staff, and Warehouse Staff. The output of the system that will be built is to assist the Production Manager in calculating optimal production requirements every day, as well as assisting in making production plan decisions for each day. © 2024 IEEE.

    Author keywords

    decision making system; optimization; production; raw material; simplex method

    This article can be accessed at https://www.scopus.com/pages/publications/105015779398

  • Comparative Analysis of Interactive Map Data Loading Using the GeoJSON Method for West Java New Entrepreneurs

    Authors: Angga Setiyadi; Eko Budi Setiawan

    DOI: 10.1109/INCITEST59455.2023.10396976

    Abstract

    The purpose of this study is to analyze interactive maps using the GeoJSON method and without using GeoJSON for new entrepreneurs in West Java. The flow of the stages of this research consists of three parts, namely analyzing the code, implementing the code, and testing the code that has been created using GeoJSON and without GeoJSON. The results obtained in this study are using GeoJSON to display data on an interactive map faster than without using GeoJSON. The downside of using GeoJSON on an interactive map is that the app developer is more likely to write down the source code on loading the interactive map. © 2023 IEEE.

    Author keywords

    Comparative Analysis; GeoJSON; Interactive Map; New Entrepreneurs

    This article can be accessed at: https://www.scopus.com/pages/publications/85185197883

  • Implementation of the K-Means Clustering Technique Using RFM Attributes to Enhance Donor Retention at Masjid Nusantara

    Authors: Iqbal Muhammad Farisi, Sri Supatmi

    DOI: 10.1109/INCITEST64888.2024.11121472

    Abstract

    The efforts aimed at donor retention are a fundamental aspect of Customer Relationship Management (CRM) strategies that enhance organizational effectiveness. This research concentrated on entities that have traditionally emphasized the acquisition of new donors over the retention of current donors, a methodology that can hinder long-term viability. By categorizing donors and formulating targeted retention strategies, this research seeks to redirect the emphases from donor acquisition to retention, by using the K -Means clustering methodology along with the RFM (Recency, Frequency, Monetary) framework. The descriptive research adopting a quantitative paradigm followed the CRISP-DM process, which included the data selection, preprocessing, transformation, processing, and analysis stages. The data analysis found four distinct clusters of donors, each of which is equipped with customized retention strategies based on their respective RFM scores. These strategies of retention contribute to pragmatic outcomes by empowering organizations to increase donor loyalty, improve donor engagement, and ultimately, facilitate their mission of building and maintaining mosques. This study demonstrates that prioritizing donor retention, which is supported by an empirically based methodology, has potential to enhance an organization’s performance and sustainable success. © 2024 IEEE.

    Author keywords

    CRISP-DM; CRM; Data Analysis; Donor Retention; K-Means Clustering; RFM

    This article can be accessed at https://www.scopus.com/pages/publications/105015849404

  • Innovative Batik NFTs Using Deep Neural Networks

    by Redefine Digital Art and Photography

    Authors: Yeffry Handoko Putra; Rahma Wahdiniwaty; Rini Maulina; Noorihan Abdul Rahman; Zuriani Ahmad Zukarnain; Wan Fariza Abdul Rahman

    DOI: 10.1109/INCITEST64888.2024.11121421

    Abstract

    Batik is an important part of Indonesia’s culture. It is known for its detailed repeating designs. Repeating designs are tough for designers, especially when they to mix in modern styles. Still, more people online are getting excited about making unique batik patterns. With new tech, like deep learning, things have changed. These models are doing better than the old ways for tasks like classifying images and recognizing objects. One cool tech, called pix2pix, helps turn one image into another using a method called Conditional Generative Adversarial Networks (cGAN). It uses data sets with images along with their edge maps that get pulled using techniques like Canny Edge Detection in OpenCV. This means that it can help create complex batik designs. This study also looks at Non-Fungible Tokens (NFTs) and how they can help sell and keep batik art safe. By turning batik designs into NFTs, artists can prove their work is accurate and connect with fans all over the world. This fresh idea not only keeps Indonesia’s rich culture alive but also lets people on the internet and batik lovers join in the creative fun. More chances exist for people to work together and explore digital art while saving their cultural heritage. Looking ahead, the Batik patterns created can become NFTs. This means that each Batik pattern’s uniqueness and ownership can be clearly shown through blockchain tech. It opens up a digital marketplace where folks can buy, sell, or collect these digital art pieces. Adding NFTs gives more value and realness to the Batik patterns made using advanced deep-learning methods. This broadens how this research could be used and its effect in the digital world. © 2024 IEEE.

    Author keywords

    batik; cultural preservation; deep learning; non-fungible tokens (NFTS)

    Indexed keywords

    Engineering controlled terms

    Arts computing; Edge detection; Electronic commerce; Historic preservation; Learning systems; Photography

    Engineering uncontrolled terms

    Adversarial networks; Batik; Cultural preservation; Data set; Deep learning; Digital art; Digital photography; Indonesia; Neural-networks; Non-fungible token

    Engineering main heading

    Deep neural networks

  • K-Means Clustering for Crime Cluster Analysis in District Prosecution Office of Indramayu

    K-Means Clustering for Crime Cluster Analysis in District Prosecution Office of Indramayu

    Author: Aditya Erlangga, Rizky Bani Asmara E.M., Raudatussholihat Amalia, Estiko Rijanto

    DOI: 10.1109/INCITEST59455.2023.10396984

    Abstract

    Crime is a criminal behavior that often causes problems in society with a variety of crimes that often occur such as murder, assault, rape, theft, fraud, embezzlement, narcotics and. In Article 35 paragraph (1) letter K of Law No. 16 of 2004 concerning the Attorney General of the Republic of Indonesia as amended in Law No. 11 of 2021 states that the Prosecution Office is a government institution whose function is related to judicial power to prosecute and other power abide by law. K-Algorithm Means is a non-hierarchical clustering algorithm that has capability to group data in large quantity, quickly and efficiently. In this paper, data from 217 cases tested from Indramayu District Attorney during January 2022 until December 2022. K-Means algorithm was successful to classify relationship between criminal cases and which prosecutors whom handling them. The result using optimal K=3 formed three cluster which consists of Cluster C1 = 9 types of cases, Cluster C2 = 9 types of cases Cluster C3 = 9 types cases with members C2 and C3 have the Global Unique, Domination, and Share characteristics within each cluster. © 2023 IEEE.

    Keywords: clustering; crime; K-Means; prosecutor

    This article can be accessed at https://www.scopus.com/pages/publications/85185198398

  • Presentation of Strategic Information in the Agriculture Domain at the Agriculture Office of Garut Regency using Data Visualization

    Authors: Muhammad Jafar Shidik; Eddy Soeryanto Soegoto; Dian Dharmayanti; Lia Warlina

    Abstract

    The purpose of this research is to produce the right visualization to help the Food Crops Sector obtain the information needed more quickly. The form of information presentation uses tables in Microsoft Excel, resulting in the field of food crops taking longer to get information. So that it hampers the understanding of information and policymaking. The method used in this research is data visualization. Based on research on data visualization, it is stated that effective data presentation is by transforming data into visual form. This research produces a visualization design and is implemented in the form of a prototype. The next stage is to test the results of the prototype using the usability testing method for food crop staff. Based on the test results, it can be concluded that the prototype of food crop data visualization can help the food crop field understand the information needed more quickly. © 2023 IEEE.

    Author keywords

    agriculture; data visualization; food crops; usabillity test; visualization

    This article can be accessed at: https://www.scopus.com/pages/publications/85185200989