Article Per Year (5 Year)
p-Index From 2020 - 2025
0.947
P-Index
This Author published in this journals
All Journal JURNAL MEDIA INFORMATIKA BUDIDARMA Jurnal Mekanova : Mekanikal, Inovasi dan Teknologi Journal Of Artificial Intelligence And Software Engineering eProceeding of TIK
Erdiansyah, Umri
Unknown Affiliation
Aerospace Engineering Automotive Engineering Computer Science & IT Control & Systems Engineering Electrical & Electronics Engineering Engineering Industrial & Manufacturing Engineering Mechanical Engineering

Published : 5 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 5 Documents
Search

 1 
Komparasi Metode K-Nearest Neighbor dan Random Forest Dalam Prediksi Akurasi Klasifikasi Pengobatan Penyakit Kutil Erdiansyah, Umri; Irmansyah Lubis, Ahmadi; Erwansyah, Kamil
JURNAL MEDIA INFORMATIKA BUDIDARMA Vol 6, No 1 (2022): Januari 2022
Publisher : STMIK Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/mib.v6i1.3373

Abstract

Warts are a skin health problem, usually characterized by small, rough bumps on the surface of the skin caused by a virus, known as the human papillomavirus (HPV). The common way of treating warts is with immunotherapy, which is the treatment of warts by strengthening the body's immune system. In the process of predicting and diagnosing warts, it can be done by applying Machine Learning. This study focuses on the comparison of the K-Nearest Neighbor classification method with Random Forest to see the level of accuracy in predicting the success of the treatment of warts. Data for Immunotheraphy was obtained from the UCI Machine Learning Repository with a total of 90 data records, 7 attributes and 1 attribute class. Based on the results of testing the K-Nearest Neighbor and Random Forest methods to see the accuracy of the prediction of the success of the data being tested, the results obtained are the accuracy of the KNN method of 90.00% and the Random Forest method with an accuracy of 85.50%. From the results obtained from the tests that have been carried out, it is known that the Random Forest method is a better method than K-Nearest Neighbor in predicting accuracy in the Immunotheraphy Dataset.
Sistem Monitoring Kehadiran Mahasiswa Politeknink Negeri Lhokseumawe Menggunakan Face Recognetion Dan Algoritma Haar Cascade Berbasis Mobile Rifki, Muhammad Aidil; Mursyidah, Mursyidah; Erdiansyah, Umri
eProceeding of TIK Vol 4, No 2 (2024): eProTIK: November, 2024
Publisher : Politeknik Negeri Lhokseumawe

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

Abstract

Penelitian ini membahas sistem pemantauan kehadiran mahasiswa di Politeknik Negeri Lhokseumawe yang menggunakanalgoritma haar cascade untuk deteksi wajah. Hasil analisis menunjukkan bahwa waktu eksekusi program meningkat secara signifikandengan bertambahnya jumlah data wajah, terutama saat data melebihi 500. Hasil analisis akurasi deteksi wajah menggunakan haarcascade menunjukkan bahwa akurasi meningkat seiring dengan bertambahnya jumlah sampel, baik pada jarak 50 cm maupun 100cm. Pada jarak 50 cm, akurasi terendah tercatat sebesar 39% dengan 100 sampel dan meningkat menjadi 62% dengan 1000 sampel,dengan rata-rata akurasi sekitar 50%. Sedangkan pada jarak 100 cm, akurasi dimulai dari 57% dengan 100 sampel dan mencapai98% dengan 1000 sampel, dengan rata-rata akurasi sekitar 71%. Perbandingan ini menunjukkan bahwa jarak 100 cm selalu lebihakurat dibandingkan dengan jarak 50 cm, yang menunjukkan pentingnya jumlah data latih yang mencukupi dan kondisi pencahayaanyang optimal untuk meningkatkan akurasi. Pengujian sistem secara keseluruhan menunjukkan performa deteksi yang baik dengansemua mahasiswa terdeteksi dan tercatat sebagai "Present". Pengujian kualitas layanan (QoS) menunjukkan bahwa throughput sistemrendah, dengan nilai 29,86 kbps dan 28,99 kbps, yang dikategorikan buruk menurut standar TIPHON, meskipun tidak ada kehilanganpaket, yang menunjukkan keandalan sistem yang sangat baik dengan indeks nilai 4. Total delay pada kedua percobaan adalah 98 msdan 105 ms, yang dikategorikan sangat baik menurut standar TIPHON dengan indeks nilai 4. Secara keseluruhan, sistem inimenunjukkan keandalan dan waktu tunda yang sangat baik, namun perlu adanya peningkatan dalam kecepatan pengiriman datauntuk memenuhi standar yang lebih tinggi.Kata kunci : Sistem Monitoring, Face recognition, Haar cascade, Deteksi Objek, Quallity of Service
Rancang Bangun Virtual Tour Reality Menggunakan Pannelum Panorama Viewer Rudi F, Fachri Yanuar; Syahputra, Guntur; Erdiansyah, Umri; bin Amirudin, Iphonsyah; Abdi, Mustainul
Jurnal Mekanova : Mekanikal, Inovasi dan Teknologi Vol 10, No 2 (2024): Oktober
Publisher : universitas teuku umar

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35308/jmkn.v10i2.10387

Abstract

Kemajuan teknologi di era digital saat ini telah membuka peluang besar dalam pengembangan aplikasi berbasis virtual reality (VR) yang dapat digunakan di berbagai sektor, termasuk pendidikan. Penelitian ini bertujuan untuk merancang dan membangun Virtual Tour Reality di instansi pendidikan yaitu gedung jurusan TIK Politeknik Negeri Lhokseumawe dan SMK N 5 Lhokseumawe dengan memanfaatkan Pannellum Panorama Viewer, sebuah pustaka JavaScript open-source untuk menampilkan gambar panorama. Dengan adanya virtual tour ini, pengguna, khususnya  calon visitor, dapat mengeksplorasi lingkungan secara interaktif tanpa harus hadir secara fisik. Pengembangan virtual tour ini dilakukan dengan metode pengumpulan data, yang meliputi pengambilan gambar panorama di berbagai titik lokasi strategis dalam kampus dan sekolah, integrasi dengan Pannellum untuk menampilkan gambar dalam bentuk panorama interaktif, serta pengujian untuk memastikan fungsionalitas dan user experience yang optimal. Hasil dari penelitian ini diharapkan dapat membantu institusi pendidikan dalam meningkatkan daya tarik dan memberikan pengalaman yang lebih imersif kepada calon visitor, terutama di era di mana mobilitas fisik terbatas.
Webcam-Based Finger Detection Using Mediapipe Rudi, Fachri Yanuar; Syahputra, Guntur; Erdiansyah, Umri; Safar, Ilham
Journal of Artificial Intelligence and Software Engineering Vol 5, No 2 (2025): June
Publisher : Politeknik Negeri Lhokseumawe

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30811/jaise.v5i2.6996

Abstract

The development of multimedia-based learning tools continues to advance alongside technological innovations, particularly in recognizing parts of the human body such as fingers. This study aims to develop an interactive learning application for identifying the names of fingers using motion capture technology through Mediapipe and the Python programming language. The application utilizes a webcam as the input device to detect hand positions and movements in real-time. The system is designed to recognize individual fingers by comparing the length of finger segments and the width of the palm, based on hand landmark data provided by the Mediapipe Hands module. Testing results show that the application achieved a 90% success rate across 50 trials. These results indicate that Mediapipe technology holds significant potential for use in the development of interactive and educational learning media.
Smart Infusion Digitalization Based on IoT, Long-Range Communication, and Cloud Ananta, Adam; Nasir, Muhammad; Erdiansyah, Umri
Journal of Artificial Intelligence and Software Engineering Vol 4, No 2 (2024)
Publisher : Politeknik Negeri Lhokseumawe

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30811/jaise.v4i2.6135

Abstract

Currently, the monitoring of infusion fluids is performed by periodically checking each patient, regardless of whether there is an obstruction or not. To address this challenge, a system based on the Internet of Things (IoT), Long Range (LoRa) at a 2.4 GHz frequency, and Cloud technology, known as the digital smart infusion system, has been developed. This system aims to enhance the efficiency and safety of infusion fluid delivery, facilitate real-time monitoring by nurses, and provide accurate and up-to-date data. The testing results indicate that the implementation of the MQTT protocol in this system yields positive outcomes, with delay times varying between 42 ms (5 minutes), 84.3 ms (10 minutes), and 73.8 ms (15 minutes), along with very low packet loss rates of 0.03% at 5 minutes, 0.02% at 10 minutes, and 0.01% at 15 minutes. Additionally, the system's throughput remains stable, with values of 92.6 Kbps at 5 minutes, 83.8 Kbps at 10 minutes, and 86.2 Kbps at 15 minutes. In tests of LoRa without obstructions, packet loss percentages remain low up to a distance of 10 meters, with a value of 0%, but then increase to 68.29% at 25 meters. Tests with obstructions show a more drastic decline in signal quality, with packet loss reaching 6.98% at 5 meters and increasing to 70.97% at 25 meters.
Co-Authors Abdi, Mustainul Ahmadi Irmansyah Lubis Ananta, Adam bin Amirudin, Iphonsyah Guntur Syahputra, Guntur Ilham Safar kamil erwansyah Muhammad Nasir Mursyidah Mursyidah Rifki, Muhammad Aidil Rudi F, Fachri Yanuar RUDI, FACHRI YANUAR