Claim Missing Document
Check
Articles

Found 29 Documents
Search

Klasifikasi Warna Objek secara Real-Time Menggunakan Optimasi Model CNN MobileNetV2 Apriliyanti, Resti; Kurniadi, Denny; Novaliendry, Dony; Rahmadika, Sandi; Farhan, Muhammad
Jurnal Teknologi Informasi dan Pendidikan Vol. 18 No. 2 (2025): Jurnal Teknologi Informasi dan Pendidikan
Publisher : Universitas Negeri Padang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24036/jtip.v18i2.969

Abstract

This research aimed to develop a Convolutional Neural Network (CNN) model for automatic object color classification using MobileNetV2. To determine the optimal configuration, the training process adjusted several hyperparameters, with particular focus on identifying the most suitable learning rate. The dataset consisted of 3,212 images grouped into five color categories: red, green, blue, random (including yellow, orange, and brown), and none (no object detected). Data augmentation techniques were applied to enhance the variety and robustness of the dataset. The model was trained using the Adam optimizer alongside the categorical crossentropy loss function, with various learning rate settings tested during training. Evaluation results showed that the model worked best with a learning rate of 0.0001 and a batch size of 32, with an average accuracy of 94%. To display prediction results in real time, the top-performing model was integrated into a graphical user interface (GUI). These findings demonstrate the effectiveness of the MobileNetV2-based CNN model in recognizing object colors and highlight its suitability for integration into real-time industrial sorting applications
A Dual-Validation Mobile Attendance Application for Discipline-Oriented Vocational Schools: Development and Evaluation Using the Waterfall Kurnia, Hasbi; Farell, Geovanne; Kurniadi, Denny; Marta, Rizkayeni
Journal of Hypermedia & Technology-Enhanced Learning Vol. 3 No. 3 (2025): Journal of Hypermedia & Technology-Enhanced Learning—Next Horizon
Publisher : Sagamedia Teknologi Nusantara

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58536/j-hytel.200

Abstract

Manual attendance systems in educational institutions, particularly in highly disciplined environments such as vocational schools with cadet programs, present significant challenges, including inaccuracy, administrative burden, and susceptibility to fraud. This study addresses these issues by designing and developing a secure mobile attendance application for the Android platform. The development process followed the Waterfall software development model, providing a structured framework from requirements analysis through testing. The proposed solution incorporates a dual-validation mechanism combining Quick Response (QR) Codes for session-specific identification with Geolocation technology to verify users’ physical presence within the designated school area. The application supports multiple user roles, including students, teachers, administrators, and parents, and offers automated attendance tracking for daily roll calls and extracurricular activities. It enables teachers to input attendance for in-class sessions manually. Empirical evaluation confirmed that the application effectively mitigates vulnerabilities inherent in the manual system. The dual-validation approach substantially enhances both the integrity and efficiency of the attendance process, offering a reliable and secure tool for monitoring student discipline. This study provides a practical model for integrating accessible digital technologies to strengthen administrative processes in secondary education contexts.
Rancang Bangun Aplikasi Pengelolaan Layanan Lansia Pada Posyandu Berbasis Web Di Puskesmas Dadok Tunggul Hitam Rahmayanti, Desy; Budayawan, Khairi; Kurniadi, Denny; Saputra, Hadi Kurnia
Journal of Innovative and Creativity Vol. 5 No. 3 (2025)
Publisher : Fakultas Ilmu Pendidikan Universitas Pahlawan Tuanku Tambusai

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

Abstract

Posyandu lansia merupakan kegiatan pelayanan kesehatan masyarakat yang berperan penting dalam pemantauan dan peningkatan kesehatan lanjut usia. Berdasarkan hasil observasi di Puskesmas Dadok Tunggul Hitam, proses pencatatan kegiatan posyandu masih dilakukan secara manual menggunakan buku register dan spreadsheet, sehingga menimbulkan kendala dalam efisiensi, keakuratan data, serta keterlambatan pelaporan. Penelitian ini bertujuan untuk merancang dan membangun aplikasi pengelolaan layanan lansia berbasis web yang dapat membantu proses pencatatan, pengelolaan data, serta pelaporan kegiatan posyandu secara terintegrasi. Metode pengembangan sistem yang digunakan adalah metode Waterfall, dengan tahapan meliputi analisis kebutuhan, perancangan, implementasi, pengujian, dan pemeliharaan. Aplikasi ini dikembangkan menggunakan framework Laravel dengan basis data MySQL dan diimplementasikan pada lingkungan berbasis web agar mudah diakses oleh kader, perawat, serta kepala puskesmas. Hasil pengujian menggunakan metode black-box testing menunjukkan bahwa seluruh fungsi sistem berjalan sesuai dengan kebutuhan pengguna. Aplikasi mampu menampilkan data lansia, hasil pemeriksaan, jadwal kegiatan, serta laporan otomatis dalam bentuk grafik dan tabel. Sistem ini meningkatkan efisiensi kerja, mengurangi kesalahan pencatatan, serta mempercepat proses pelaporan ke pihak puskesmas. Dengan demikian, aplikasi ini berhasil mendukung transformasi digital dalam pelayanan kesehatan masyarakat khususnya pada kegiatan posyandu lansia di Puskesmas Dadok Tunggul Hitam
Implementasi Teknologi Blockchain Pada Aplikasi E-Voting Berbasis WEB Ilahi, Akbar; Kurniadi, Denny; Novaliendry, Dony; Sriwahyuni, Titi
Jurnal Pendidikan Tambusai Vol. 8 No. 2 (2024)
Publisher : LPPM Universitas Pahlawan Tuanku Tambusai, Riau, Indonesia

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

Abstract

E-voting adalah sistem pemungutan suara elektronik yang memanfaatkan teknologi seperti komputer atau perangkat seluler untuk mempermudah proses pemilihan dan mengurangi risiko kecurangan serta kesalahan manusia. Dengan e-voting, proses perhitungan suara lebih cepat, biaya pencetakan surat suara bisa dihemat, dan peralatan dapat digunakan berulang kali. Namun, e-voting memiliki kelemahan dalam hal keamanan, seperti potensi perubahan surat suara dan manipulasi data pemilih. Blockchain, khususnya Ethereum local blockchain Ganache, dapat mengatasi beberapa tantangan ini dengan meningkatkan keamanan data pemilihan. Penggunaan model prototype dalam pengembangan e-voting membantu memperjelas spesifikasi dan kebutuhan sistem.
Integrasi Metode Analytic Hierarchy Process (AHP) dan Sistem Digital Untuk Penilaian Kompetensi Mahasiswa Informatika Deanof Putri, Feni; Kurniadi, Denny
Jurnal Pendidikan Tambusai Vol. 9 No. 2 (2025): Agustus
Publisher : LPPM Universitas Pahlawan Tuanku Tambusai, Riau, Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31004/jptam.v9i2.31134

Abstract

Penerapan metode Analytic Hierarchy Process (AHP) bertujuan untuk meningkatkan objektivitas dan efisiensi penilaian kompetensi mahasiswa, khususnya dalam proses observasi. Penelitian ini mengimplementasikan AHP pada sistem penilaian berbasis web untuk mendukung uji kompetensi di Program Studi Pendidikan Teknik Informatika. Metode penelitian yang digunakan adalah pendekatan rekayasa perangkat lunak model Waterfall, dengan pengumpulan data melalui observasi dan studi pustaka menggunakan data dummy untuk menguji algoritma. Hasilnya menunjukkan AHP berhasil memberikan bobot prioritas yang proporsional pada kriteria penilaian, sangat penting (0,653), penting (0,251), dan pendukung (0,096). Perhitungan ini divalidasi dengan nilai rasio konsistensi (CR) sebesar 0,017, yang menunjukkan matriks konsisten. Implementasi sistem ini mampu mengurangi bias subjektivitas, meningkatkan transparansi, dan memberikan hasil secara real-time.
ANALISIS KEPUASAN PENGGUNA APLIKASI AUTOMATIC METER READING (AMR) MENGGUNAKAN METODE UTAUT Fitra, Anggi; Kurniadi, Denny
INCARE, International Journal of Educational Resources Vol. 1 No. 4 (2020): December 2020
Publisher : FKDP (Forum Komunikasi Dosen Peneliti)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59689/incare.v1i4.140

Abstract

This study aims to reveal the level of trust in the use of the AMR (Automatic Meter Reading) application implemented at PLN. The facts of satisfaction include performance expectancy, effort expectancy, social influence, facilitating conditions, and attitude. The research method used is descriptive research through the use of the UTAUT model. The purpose of this research is to determine the level of satisfaction in using AMR applications using the UTAUT method. The results of the study found that (1) the variable performance expectancy contributed 8.1%, obtained from the percentage contribution of the performance expectancy variable to the Behavioral Intention to use system in the acceptance and use of the Automatic Meter Reading (AMR) system (2) the effort expectancy variable contributed 13, 6%, obtained from the percentage contribution of the effort expectancy variable to use the system in the acceptance and use of the Automatic Meter Reading (AMR) system (3) the social influence variable contributes 21.16%, obtained from the percentage contribution of the social influence variable to the Behavioral Intention to use system In the acceptance and use of the Automatic Meter Reading (AMR) system (4) the facilitating conditions variable contributes 23.52%, the results are the percentage of contribution of the facilitating conditions variable to the intention of use behavior towards the Behavioral Intention to use system (BIUS) and (5) the variable Attitude contributed 10.43%, obtained didap from the percentage contribution of the Attitude variable to the Behavioral Intention to use system in the acceptance and use of the Automatic Meter Reading (AMR) system.
IMPLEMENTASI BLOCKCHAIN DALAM APLIKASI PEMILU Pratama, Yulfitno Wingga; Kurniadi, Denny
INCARE, International Journal of Educational Resources Vol. 2 No. 2 (2021): August 2021
Publisher : FKDP (Forum Komunikasi Dosen Peneliti)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59689/incare.v2i2.256

Abstract

Elections are a democratic activity in Indonesia which are carried out every five years to elect leaders, but in the process of manual election calculations, the results are leaked and can cause data security problems. Through this research, we can provide solutions by implementing blockchain technology in election applications that will provide security for election results data stored on the blockchain. Blockchain technology is a chain of data blocks that are connected to each other by peer to peer. This application uses the Solidity programming language and uses a local ethereum blockchain database (Ganache). The method used in making this application is the waterfall method with Unified Modeling Language (UML) modeling where the visual modeling method is in object-oriented system design. From the results of testing the blockchain system for this application, it is found that blockchain can help secure election results very securely, where every result data will be stored in every block in a decentralized blockchain network.
The application of naive bayes method for final project topic selection within the project-based learning framework in the data mining course Kurniadi, Denny
Jurnal EDUCATIO: Jurnal Pendidikan Indonesia Vol. 10 No. 1 (2024): Jurnal EDUCATIO: Jurnal Pendidikan Indonesia
Publisher : Indonesian Institute for Counseling, Education and Therapy (IICET)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29210/1202423794

Abstract

This study aims to explore the potential application of the Naive Bayes Method in predicting the determination of final project topics for students within the context of project-based learning (PBL) in the Data Mining course. Adopting an observational quantitative approach, the research involved 34 students from the Educational Technology Informatics Study Program at Padang State University between 2013-2015 as research subjects. The sampling technique employed was stratified random sampling, aligning with the research findings. The study's results indicate that the Naive Bayes Method is proficient in providing accurate predictions concerning the determination of final project topics, with significant accuracy, precision, and recall values. Moreover, the integration with project-based learning (PjBL) approach effectively enhances the authenticity and relevance of the learning experience for students, further substantiating the effectiveness of Project-Based Learning (PjBL).
Application of data mining for analysis of consumer purchase data on sales transaction data at Halal Mart Hni Hpai Dharmasraya Yuliarnis, Sri Kurnia; Hendriyani, Yeka; Kurniadi, Denny; Giatman, M
Jurnal Pendidikan Teknologi Kejuruan Vol 3 No 1 (2020): Reguler Issue
Publisher : Universitas Negeri Padang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24036/jptk.v3i1.6923

Abstract

The sales strategy determines the continuity of the business being run. The problems that occur are the sales archive data has not been analyzed in-depth, the information system has not been integrated with applications for sales data analysis, online media promotion has not been maximized, inadequate stock of goods, the layout of goods is not optimal, and the combination of the number of products is not optimal. This study aims to extract hidden information in the sales database using Data Mining. From the information generated, sales strategy recommendations are developed relating to promotions, inventory, catalogue design, item layout, and the combination of product quantities. The method used is the association rule with Apriori algorithm to find consumer purchase patterns through the resulting association. The importance of association can be identified by two benchmarks, namely support and confidence. The sales strategy analyzed includes product promotion, catalogue design, product layout, stock predictions, and product combinations for sale. Based on the research produced 7 strong rules which are the highest association rules which are then developed into a sales strategy recommendation.