Garuda - Garba Rujukan Digital

Optimasi Algoritma K-Nearest Neighbors Menggunakan GridSearchCV untuk Klasifikasi Penyakit Diabetes Yaqin, Ainul; Kurniawan, Defri; Zeniarja, Junta
Infotekmesin Vol 16 No 1 (2025): Infotekmesin: Januari 2025
Publisher : P3M Politeknik Negeri Cilacap

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35970/infotekmesin.v16i1.2557

Diabetes is a chronic disease that has a significant impact on global health, with prevalence increasing every year. Therefore, early detection is crucial to prevent further complications and save lives. The utilization of technology, such as machine learning, offers innovative solutions to improve the accuracy of predicting this disease. This research develops a diabetes prediction model using the K-Nearest Neighbors (KNN) algorithm with the Pima Indians Diabetes Database dataset. Given the class imbalance in the dataset, Random Over-Sampling technique was applied to balance the data distribution. The results showed that the KNN model optimized with GridSearchCV resulted in 88% accuracy, 89% precision, 75% recall, and 82% F1-score. This approach is expected to produce a more accurate and efficient model to support early detection of diabetes, and shows the great potential of machine learning technology in improving the effectiveness of disease diagnosis and control.

Pendampingan Pemanfaatan Google Site Sebagai Media Pembelajaran Berbasis Web di SMPN 7 Semarang Rakasiwi, Sindhu; Kurniawan, Defri; Hidayat, Erwin Yudi; Zeniarja, Junta; Dzaky, Azmi Abiyyu; Haresta, Alif Agsakli
ABDIMASKU : JURNAL PENGABDIAN MASYARAKAT Vol 8, No 2 (2025): MEI 2025
Publisher : LPPM UNIVERSITAS DIAN NUSWANTORO

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.62411/ja.v8i2.2970

A website is the heart of an institution, school or company profile. With a web appearance that is always active and always has useful content, it will add to the image of the owner of the website. Because of this, the community service team wants to provide assistance to teachers so that they can also contribute to filling the website. So not only IT teachers can contribute to the website, but all teachers can contribute so that the website can be more active and interactive for students, parents of students and even for the general public who want to know information about SMPN 07 Semarang. And through this assistance, it also utilizes the Google site for more interactive learning and students are also more active in creating learning for the future.

Collaborative Governance dalam Pengelolaan Badan Usaha Milik Desa (Bumdes) “Cipta Karya Unggul” di Desa Tipar Kidul Kecamatan Ajibarang Kabupaten Banyumas Kurniawan, Defri; Amartha, Putri; Azizah, Hidayatul; Fauziah, Mitha
Indonesian Journal of Public Administration Review Vol. 3 No. 1 (2025): November
Publisher : Indonesian Journal Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47134/par.v3i1.5042

Village-Owned Enterprises (BUMDes) play an essential role in improving rural economic development through the management of local potential. This study aims to identify the process of Collaborative Governance in the management of BUMDes “Cipta Karya Unggul” in Tipar Kidul Village, Ajibarang District, Banyumas Regency. This research uses a descriptive qualitative approach with data collected through interviews, observation, and documentation, while informants were selected using purposive sampling. The findings indicate that the management of BUMDes Cipta Karya Unggul involves collaboration among the village government, the community, and the private sector, particularly PT. Sinar Tambang Artha Lestari (PT STAR). This collaboration reflects the principles of Collaborative Governance—participatory, transparent, and goal-oriented— resulting in increased village income and community welfare. The success of this collaboration is supported by facilitative leadership, open communication, and mutual trust among stakeholders. Therefore, the application of the Collaborative Governance model is an effective strategy for strengthening sustainable village economic governance.

Rancang Bangun Sistem Try Out Berbasis Paperless untuk Evaluasi Hasil Belajar Siswa dengan MVC Sani, Ramadhan Rakhmat; Kurniawan, Defri
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 6 No 3: Juni 2019
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Sistem pelaksanaan Ujian Nasional di Indonesia saat ini mulai beralih menggunakan komputer sebagai media dalam pelaksanaannya menggantikan sistem yang lama. Sebagai bentuk dukungan dalam mengurangi penggunaan kertas dalam pembelajaran dan evaluasi baik lembar soal maupun materi di sekolah. Tujuan dari penelitian ini adalah menghasilkan purwarupa model sistem tryout berbasis paperless untuk evaluasi hasil belajar siswa untuk menggantikan pemakaian kertas dengan konsep yang diberikan oleh media komputer. Sehingga memberikan keuntungan dalam efisiensi waktu dan biaya, mengurangi kecurangan serta mempercepat dalam proses evaluasi. Metode yang digunakan dalam pengembangan aplikasi ini menggunakan Rapid Application Development (RAD) yang meliputi tahapan analisa kebutuhan perangkat lunak, perancangan perangkat lunak, implementasi perangkat lunak dan pengujian perangkat lunak dengan penerapan konsep Sistem Development Life Cycle (SDLC) sehingga cepat untuk dievaluasi oleh pengguna. Untuk bahasa pemodelan sistem menggunakan UML (Unified Modeling Language) yang terdiri use case diagram, sequential diagram dan pemodelan database. Penggunaan framework Codeigniter memberikan kemudahan dalam konsep Object Oriented Programing (OOP) dengan menggunakan arsitektur web MVC (Model, View, Controller). Pemisahan logika pembuatan kode pada tampilan website menjadikan lebih terstruktur, sederhana dan aman sehingga memberikan kemudahan bagi developer maupun programmer dalam pengembangannya tanpa harus dimulai dari awal. Hasil pengujian sistem menggunakan blackbox testing menunjukan hasil yang baik dan sudah mencapai 90% dari prinsip usability yang telah diimplementasikan.AbstractThe current National Examination System in Indonesia has begun to switch to using computers as a medium in its implementation to replace the old system. As a form of support in reducing paper use in learning and evaluating both question sheets and material at school. The purpose of this study was to produce a prototype paperless based tryout system model for evaluating student learning outcomes to replace paper use with the concepts provided by computer media. So as to provide benefits in time and cost efficiency, reduce fraud and accelerate the evaluation process. The method used in the development of this application uses Rapid Application Development (RAD) which includes the stages of software requirements analysis, software design, software implementation and software testing with the application of the concept of System Development Life Cycle (SDLC) so that it is quickly evaluated by users. For system modeling languages use UML (Unified Modeling Language) which consists of use case diagrams, sequential diagrams and database modeling. The use of CodeIgniter framework provides convenience in Object Oriented Programing (OOP) concepts using the MVC web architecture (Model, View, Controller). Separation of logic in making code on the website display makes it more structured, simpler and safer so that it makes it easy for developers and programmers to develop without having to start from scratch. The results of testing the system using blackbox testing showed good results and has reached 90% of the usability principle that has been implemented.

Pengembangan Sistem Modul Komisi Dinamis pada Modul Penjualan ERP - Odoo12 Wahyu Utomo, Danang; Kurniawan, Defri; Rosi Subhiyakto, Egia
Infotekmesin Vol 12 No 2 (2021): Infotekmesin: Juli 2021
Publisher : P3M Politeknik Negeri Cilacap

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35970/infotekmesin.v12i2.729

The improvement of the sales system not only focuses on the advantage result of the sales transaction but also can use another parameter to improve it. One of a parameter used is commission. Giving commissions to the salesperson can improve their work performance and have an impact on increasing sales targets. Based on the study literature, the problem faced by the company is the discrepancy of commission. It canbe affected by several factors such as the commission system are not integrated with the main system, improper formula, or there are many systems used in the company so it the staff are difficult to integrate the system. For example, the company using Odoo ERP to support sales transaction and use commission information system separately. The salesperson must integrate sales data into both of the systems. It can affect the time delay of decision commission. Based on the problem above, we propose a prototype commission system that integrates with Odoo12. The salesperson does not need to integrate data manually into the system because it automatically integrates into the system. This study uses a prototyping model as a software development method. The results show that the commission system can implement on the Odoo12 ERP to decide commission to the salesperson. 70% of respondent agree that system has able to use in order to setting up commission module on Odoo

Optimasi Algoritma K-Nearest Neighbors Menggunakan GridSearchCV untuk Klasifikasi Penyakit Diabetes Yaqin, Ainul; Kurniawan, Defri; Zeniarja, Junta
Infotekmesin Vol 16 No 1 (2025): Infotekmesin: Januari 2025
Publisher : P3M Politeknik Negeri Cilacap

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35970/infotekmesin.v16i1.2557

Diabetes is a chronic disease that has a significant impact on global health, with prevalence increasing every year. Therefore, early detection is crucial to prevent further complications and save lives. The utilization of technology, such as machine learning, offers innovative solutions to improve the accuracy of predicting this disease. This research develops a diabetes prediction model using the K-Nearest Neighbors (KNN) algorithm with the Pima Indians Diabetes Database dataset. Given the class imbalance in the dataset, Random Over-Sampling technique was applied to balance the data distribution. The results showed that the KNN model optimized with GridSearchCV resulted in 88% accuracy, 89% precision, 75% recall, and 82% F1-score. This approach is expected to produce a more accurate and efficient model to support early detection of diabetes, and shows the great potential of machine learning technology in improving the effectiveness of disease diagnosis and control.

Pendampingan Pembuatan Media Pembelajaran Berbasis Multimedia Bagi Guru SD Negeri Pedurungan Kidul 02 Semarang Utomo, Danang Wahyu; Kartikadarma, Etika; Dolphina, Erlin; Kurniawan, Defri; Purwanto, Purwanto
ABDIMASKU : JURNAL PENGABDIAN MASYARAKAT Vol 7, No 1 (2024): JANUARI 2024
Publisher : LPPM UNIVERSITAS DIAN NUSWANTORO

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.62411/ja.v7i1.1802

Media pembelajaran saat ini telah berkembang, salah satu contohnya adalah media pembelajaran digital atau biasa disebut literasi digital. Literasi digital dapat berupa teks, audio, atau video. Cara mendapatkannya dapat melalui berbagai sumber seperti media sosial dan halaman web. Keuntungan dari media pembelajaran digital adalah dapat meningkatkan kemampuan belajar siswa. Guru juga dapat menggunakan berbagai sumber seperti teks, gambar, audio dan video dalam materi pembelajaran. Maka program kemitraan Masyarakat (PKM) dari Udinus menawarkan pendampingan pembuatan media pembelajaran berbasis multimedia dengan Canva. Metode yang digunakan dalam program kemitraan Masyarakat adalah praktek dengan Canva. Dalam praktek tersebut, para guru diawali membuat slide presentasi kemudian diubah menjadi video pembelajaran dengan memanfaatkan asset yang disediakan oleh Canva.

Analisis Sentimen Ulasan Mobile JKN pada Playstore dengan Perbandingan Akurasi Algoritma Naïve Bayes dan SVM Pranata, Eka Arya; Budiman, Fikri; Kurniawan, Defri
Building of Informatics, Technology and Science (BITS) Vol 7 No 1 (2025): June (2025)
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i1.7334

The facilities provided by BPJS Health by releasing the Mobile JKN application, with this application the administrative process that previously had to be done directly can be done online and more flexibly. This research aims to see the sentiment of the community towards the JKN Mobile application review by comparing the SVM and Naïve Bayes algorithms. As well as optimizing the Naïve Bayes algorithm by using grid search. Reviews are taken from Google play with the help of Google Play Scraper API, the dataset taken amounted to 7,000 reviews. The results of using Naïve Bayes with an accuracy value of 86%, after tuning optimization using Grid Search significantly increases the accuracy value of the Naïve Bayes algorithm to 91% and for the SVM algorithm has an accuracy value of 92%. From the trial, it was found that the SVM algorithm is still better than the Naïve Bayes algorithm even though it has been optimized, but by optimizing the accuracy value Naïve Bayes is closer to SVM performance. This research can provide insight into the comparison of the two algorithms in identifying JKN Mobile reviews and the need for optimization to improve the performance of algorithms in sentiment analysis, besides that this research also contributes to the improvement and development of the JKN Mobile application so that it is useful for the community.

Analisis Sentimen Pengguna X terhadap Kasus Korupsi Gula Tom Lembong Menggunakan Naïve Bayes, SVM, dan Random Forest Kuncoro, Aneira Vicentiya; Budiman, Fikri; Kurniawan, Defri
Building of Informatics, Technology and Science (BITS) Vol 7 No 3 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i3.8577

The alleged sugar import corruption case involving Tom Lembong has become one of the most widely discussed public issues on social media, generating diverse reactions. This phenomenon illustrates how public opinion on legal issues is often influenced by perceptions of the public figures involved. This study aims to analyze public sentiment regarding the case on the social media platform X (formerly Twitter). The dataset consists of 1,802 tweets collected through a crawling process using the X API with the keyword “Tom Lembong.” The research stages include data cleaning, case folding, text normalization, tokenizing, stopword removal, stemming, sentiment labeling using a lexicon-based approach, and feature extraction with the Term Frequency–Inverse Document Frequency (TF-IDF) method. The prepared dataset was then tested using three classification algorithms: Naïve Bayes, Support Vector Machine (SVM), and Random Forest. The results show that the SVM algorithm achieved the highest accuracy (84%), followed by Random Forest (80%) and Naïve Bayes (76%). Based on the sentiment labeling results, positive sentiment dominated with 61%, while negative sentiment accounted for 39%. Although the analyzed issue concerns an alleged corruption case, the dominance of positive sentiment indicates that public opinion tends to focus on Tom Lembong’s personal image or public track record, which is viewed positively rather than on the substance of the legal allegations. These findings demonstrate the effectiveness of the SVM algorithm in analyzing high-dimensional text and provide insights into how public perception of legal issues can be influenced by image factors and the socio-political context on social media.

Model Klasifikasi Cerdas Gangguan Tidur Berbasis Machine Learning Random Forest pada Data Kesehatan dan Perilaku Harian Ni'mah, Laila Maulin; Kurniawan, Defri
Building of Informatics, Technology and Science (BITS) Vol 7 No 3 (2025): December 2025
Publisher : Forum Kerjasama Pendidikan Tinggi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/bits.v7i3.8631

Sleep disorders, such as insomnia and sleep apnea, have become a significant health issue in the modern era, driven by the demands of lifestyle changes. This condition highlights the urgent need for early detection tools that are not only accurate but also easily accessible to the general public. This research aims to design and implement an intelligent classification system to automatically identify the risk of sleep disorders based on health and daily behavior data. To achieve this goal, this study applies a machine learning method using the Random Forest algorithm, which was chosen for its reliable ability to handle complex and non-linear data relationships. The data used is the "Sleep Health and Lifestyle Dataset" sourced from the Kaggle platform, covering 374 respondents with 13 relevant features. The research process included data pre-processing steps to ensure input quality, model training, and rigorous performance evaluation. The evaluation results on the test data show that the developed Random Forest model exhibited very solid performance, successfully achieving an accuracy rate of 91% and a weighted average F1-Score of 0.90. This F1-Score metric, which balances precision and recall, confirms that the model is not only accurate but also has a balanced performance in detecting each class, which is crucial for health classification. Furthermore, the feature importance analysis confirmed that Stress Level, BMI Category, and Heart Rate are the three most dominant predictor factors. The culmination of this research is the successful implementation of this predictive model into an interactive web application developed using the Streamlit framework. This application allows users to independently input their health data and receive feedback in the form of a real-time risk prediction. With an intuitive interface and easy-to-understand results, this application serves as a practical and informative initial screening tool for personal sleep health analysis.

Title

Found 21 Documents
Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 21 Documents Search

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 21 Documents
Search