Claim Missing Document
Check
Articles

Klasifikasi Presiden Republik Indonesia Menggunakan SVM Kernel Polynomial Dengan Fitur Ektraksi Gabor Kristianingrum, Kristianingrum; Rahman, Aviv Yuniar; Istiadi, Istiadi
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 9 No 7: Spesial Issue Seminar Nasional Teknologi dan Rekayasa Informasi (SENTRIN) 2022
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.2022976748

Abstract

Indonesia adalah negara dengan sistem demokrasi dalam pemerintahannya. Adanya pemilihan presiden yang dilakukan selama 5 tahun sekali dari masa kemerdekaan sampai dengan sekarang. Pemilihan presiden atau yang sering disebut dengan pemilu (pemilihan umum) ini berguna untuk memilih calon presiden dan wakil presiden dalam sebuah negara. Mengingat adanya pergantian presiden setelah 5 tahun dalam 2 periode, para remaja jaman sekarang cenderung mengikuti jaman millennial. Sehingga banyak diantaranya tidak mengenali siapa saja presiden-presiden yang pernah menjabat di Indonesia. Oleh karena itu peneliti mengusulkan Sistem Klasifikasi Presiden Republik Indonesia menggunakan SVM Kernel Polynomial dengan Fitur Ekstraksi Gabor. Tujuan dalam peneliti ini untuk membedakan dan mengklasifikasikan nama presiden berdasarkan dengan foto tersebut. Hasil dalam SVM fitur Gabor kernel Polynomial mendapatkan nilai accuracy tertinggi sebesar 80.77 dengan split ratio 10:90. Parameter precision memiliki nilai tertinggi mencapai 32.56 dengan split ratio 10:90 dan Recall mencapai 32.70 pada split ratio 10:90. Hasil dalam pengujian ini menunjukkan bahwa SVM fitur Gabor kernel Polynomial ialah yang mampu mengklasifikasikan foto presiden dengan baik dan akurat. AbstractIndonesia is a country with a democratic system in its government. Presidential elections are held every 5 years from the time of independence until now. Presidential elections or what are often called elections (general elections) are useful for selecting presidential and vice presidential candidates in a country. considering the change of president after 5 years in 2 periods, today's youth tend to follow the millennial era. So many of them do not know who the presidents who have been in Indonesia are. Therefore, the researcher proposes the Classification System for the President of the Republic of Indonesia using SVM Kernel Polynomial with Gabor Extraction Features. The purpose of this research is to distinguish and classify the name of the president based on the photo. The results in the SVM Gabor Polynomial kernel feature get the highest accuracy value of 80.77 with a split ratio of 10:90. The precision parameter has the highest value reaching 32.56 with a split ratio of 10:90 and Recall reaching 32.70 at a split ratio of 10:90. The results in this test show that SVM features a Gabor Polynomial kernel which is able to classify presidential photos well and accurately.
LOBSTER AGE DETECTION USING DIGITAL VIDEO-BASED YOLO V8 ALGORITHM Nusman, Bayu; Rahman, Aviv Yuniar; Putera, Rangga Pahlevi
Jurnal Teknik Informatika (Jutif) Vol. 5 No. 4 (2024): JUTIF Volume 5, Number 4, August 2024
Publisher : Informatika, Universitas Jenderal Soedirman

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.52436/1.jutif.2024.5.4.2144

Abstract

Lobster is an aquatic animal that has high economic value in the fishing industry. Demand for lobster in both domestic and export markets continues to increase thanks to its delicious meat and a variety of desirable dishes. Indonesia, especially Java Island, contributes significantly to the national lobster production. However, the current manual determination of lobster age has limitations such as complexity, time required, and subjectivity in assessment.To overcome this problem, this research proposes the detection of lobster age using the YOLO (You Only Look Once) method, specifically the YOLOv8 version. This algorithm is known to be able to perform image and video recognition quickly and produce high accuracy. YOLOv8 can be run using a GPU, enabling parallel operations that significantly increase the speed of object detection compared to using a CPU alone. The data processing in this study involves several stages, starting from pre-processing in the form of image extraction and bounding from lobster videos. Next, the YOLOv8 algorithm was used to train the model with customized grid and bounding box parameters. The trained model is then validated and tested using lobster image and video data. The results of the test show that the developed YOLOv8 model has a precision of 0.997, recall of 0.998, mAP50 of 0.995, and mAP50-95 of 0.971. This shows that the model is able to detect and determine the age of the lobster with very high accuracy, providing a more efficient and objective solution than the manual method.
KLASIFIKASI TINGKAT TUTUR BAHASA SASAK BERBASIS TEKS MENGGUNAKAN NAIVE BAYES arrozi, lalu muhamad alawi; rahman, aviv yuniar; putra, rangga pahlevi
Jurnal Informatika dan Teknik Elektro Terapan Vol. 12 No. 3S1 (2024)
Publisher : Universitas Lampung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.23960/jitet.v12i3S1.5149

Abstract

Abstrak. Tujuan dari penelitian ini adalah untuk menilai efektivitas klasifikasi tingkat tutur bahasa Sasak berbasis teks menggunakan algoritma Naive Bayes dalam mengidentifikasi dan mengkategorikan tingkat tutur bahasa Sasak. Penurunan kesadaran kaum muda mengenai penggunaan "tatakrama" atau tingkat tutur dalam percakapan sehari-hari di Lombok menunjukkan perlunya melestarikan aspek budaya yang penting ini. Tingkat tutur, yang melibatkan sistem kode untuk menyampaikan kesopanan, mencakup kosakata dan aturan leksikal tertentu. Metode yang digunakan dalam penelitian ini adalah Naive Bayes, yang memanfaatkan probabilitas dan statistik untuk klasifikasi teks. Ada dua tahap utama dalam studi ini: pelatihan dan pengujian, dengan pembagian data 70:30. Temuan menunjukkan bahwa model Naive Bayes mencapai F1-score sebesar 84,99%, akurasi 85,08%, presisi 85,12%, dan recall 85,08%. Hasil ini menunjukkan bahwa Naive Bayes adalah metode yang efektif untuk mengklasifikasikan tingkat tutur bahasa Sasak, meskipun hasilnya tidak setinggi beberapa studi sebelumnya. Penelitian ini memberikan kontribusi terhadap pengembangan metode yang lebih efisien dan akurat untuk klasifikasi teks tingkat tutur bahasa Sasak dan menunjukkan perlunya perbaikan dalam pemilihan fitur serta perluasan dataset untuk studi-studi mendatang.
ANALISIS SENTIMEN APLIKASI SHOPEE, TOKOPEDIA, LAZADA DAN BLIBLI MENGGUNAKAN LEKSIKON DAN RANDOM FOREST Syah, Adryan; Nurdiyansyah, Firman; Rahman, Aviv Yuniar
Jurnal Informatika dan Teknik Elektro Terapan Vol. 12 No. 3S1 (2024)
Publisher : Universitas Lampung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.23960/jitet.v12i3S1.5155

Abstract

Abstrak. Dalam era digital, aplikasi e-commerce telah menjadi sarana utama bagi masyarakat untuk berbelanja. Keberhasilan aplikasi e-commerce tidak hanya bergantung pada fungsionalitasnya tetapi juga pada pengalaman pengguna. Ulasan pengguna di Play Store menjadi indikator penting dalam mengevaluasi kepuasan dan sentimen pengguna terhadap aplikasi tertentu. Penelitian ini bertujuan untuk menganalisis sentimen pada ulasan aplikasi Shopee, Tokopedia, Lazada, dan Blibli di Play Store menggunakan pendekatan Lexicon-based dan algoritma Random Forest. Metode ini dipilih untuk memberikan interpretasi yang jelas terhadap sentimen teks dan meningkatkan akurasi analisis sentimen. Hasil penelitian menunjukkan bahwa aplikasi Lazada memiliki kinerja terbaik dengan akurasi 88,33%, presisi 88,88%, recall 88,33%, dan F1 score 88,34%. Aplikasi Blibli berada di posisi kedua dengan akurasi 85,66%, presisi 85,82%, recall 85,66%, dan F1 score 85,60%. Shopee memiliki akurasi 85,16%, presisi 85,62%, recall 85,16%, dan F1 score 85,26%. Tokopedia menunjukkan performa terendah dengan akurasi 80,33%, presisi 80,96%, recall 80,33%, dan F1 score 80,12%. Penelitian ini menunjukkan bahwa rasio pembagian data latih dan data uji mempengaruhi kinerja model, dengan model bekerja lebih efektif ketika jumlah data latih lebih besar dari data uji.
PERBANDINGAN ALGORITMA NAIVE BAYES DAN KNN DALAM ANALISIS SENTIMEN ULASAN PENGGUNA APLIKASI CAPCUT Muslim, Shinta Nilam Sari; Nurdiyansyah, Firman; Rahman, Aviv Yuniar
Jurnal Informatika dan Teknik Elektro Terapan Vol. 12 No. 3S1 (2024)
Publisher : Universitas Lampung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.23960/jitet.v12i3S1.5156

Abstract

Abstrak. Penelitian ini berfokus pada analisis sentimen ulasan pengguna aplikasi CapCut yang tersedia di Google Play Store dengan menerapkan model Naïve Bayes dan K-Nearest Neighbors (KNN). Tujuan utama penelitian ini adalah untuk mengevaluasi bagaimana pengruh variasi rasio pembagian data latih dan uji terhadap kinerja kedua metode dalam analisis sentimen, serta membandingkan keduanya berdasarkan akurasi, presisi, recall dan f1 score. Menggunakan sembilan rasio pembagian data, ditemukan bahwa rasio 80:20 memberikan kinerja terbaik untuk kedua metode. Naïve Bayes mengungguli KNN dengan akurasi 79.41% dibanding 75.63%. Rasio 50:50 memberikan presisi terbaik untuk kedua metode. Secara keseluruhan, Naïve Bayes menunjukkan performa lebih baik, terutama pada rasio 80:20, menjadikannya pilihan yang lebih tepat untuk analisis sentimen aplikasi CapCut.
Comparison of Machine Learning as an Inference Engine to Improve Expert Systems in Dengue Disease Istiadi, -; Marisa, Fitri; Joegijantoro, Rudy; Suksmawati, Affi Nizar; Rahman, Aviv Yuniar
JOIV : International Journal on Informatics Visualization Vol 9, No 3 (2025)
Publisher : Society of Visual Informatics

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.62527/joiv.9.3.3192

Abstract

Dengue disease remains a significant public health challenge in tropical and subtropical regions, with rising incidence and mortality rates over the past few decades. While expert systems have been developed for early detection, traditional approaches often rely on rigid rule-based inference engines, which are limited by their dependence on expert-defined structures and lack adaptability to evolving knowledge sources. This study introduces a novel approach to enhance the flexibility and adaptability of expert systems by integrating machine learning (ML) techniques into the inference engine, leveraging the growing availability of medical record data as a dynamic knowledge source. Using a dataset of 90 medical records, balanced to 126 items via the Synthetic Minority Over-sampling Technique (SMOTE), we evaluated the performance of multiple ML algorithms, including Decision Trees (DT), Support Vector Machines (SVM), and Artificial Neural Networks (ANN), against traditional models like Naive Bayes (NB) and K-Nearest Neighbors (KNN). The DT, SVM, and ANN models demonstrated exceptional performance, achieving average accuracy, precision, recall, and F1 scores of 97.73%, 98.33%, 97.22%, and 97.41%, respectively. The key innovation of this research lies in developing an adaptive inference engine that can dynamically learn from medical data, reducing reliance on static rule bases and enabling the expert system to evolve with new knowledge. This approach improves diagnostic accuracy and provides a scalable and flexible framework for addressing other infectious diseases. By bridging the gap between expert systems and machine learning, this study paves the way for more intelligent, data-driven healthcare solutions with significant implications for public health and disease management.
Determination of Training Participants in Community Work Training Centers Using the Naïve Bayes Classifier Algorithm Hananto, April Lia; Hananto, Agustia; Huda, Baenil; Rahman, Aviv Yuniar; Novalia, Elfina; Priyatna, Bayu
JOIV : International Journal on Informatics Visualization Vol 8, No 3 (2024)
Publisher : Society of Visual Informatics

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.62527/joiv.8.3.1995

Abstract

Community work training centers are skills training institutions that aim to improve the skills of the surrounding community by providing training programs that align with industry needs. Registration of training participants at the Al-Ikhwan Islamic Boarding School community work training centers often faces obstacles, namely, the selection process is still manual, so it takes a long time, and there is a possibility of errors. This study aims to apply the Naive Bayes Classifier Algorithm to determine whether applicants pass training at the Al-Ikhwan Islamic Boarding School community work training centers. This classification method is used to help optimize the applicant selection process by considering administrative factors, income, and training quotas. RapidMiner software is used as a tool to implement the algorithm. This study found that the Naive Bayes Classifier Algorithm can provide good accuracy results in determining applicants who pass the training selection. The test results show that the resulting model has an accuracy of 90.00% in determining passing training participants with data that has the highest chance of passing, namely data that has the attributes of the female gender, age 20 years, last education Senior High School/Vocational High School, student work/student, income 364,912, father's work as laborer, father's income 3912,280, mother's work as an IRT, and mother's income 885,964. This research increases efficiency and accuracy in determining training applicants at the Al-Ikhwan Islamic Boarding School community work training centers.
DETECTION OF LIKURAI DANCE MOVEMENT TYPES IN MALAKA REGENCY USING YOLOV8 BASED ON VIDEO Da Costa, Zania Abuk; Rahman, Aviv Yuniar; Putra, Rangga Pahlevi
JIKO (Jurnal Informatika dan Komputer) Vol 7, No 3 (2024)
Publisher : Universitas Khairun

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33387/jiko.v7i3.8815

Abstract

Indonesia is rich in traditional dances from every region, including the Likurai Dance, originating from East Nusa Tenggara, specifically in Malaka and Belu districts. This dance carries deep symbolic and historical meaning; however, it is currently threatened by lifestyle changes and globalization. Despite this, accurately and in real-time recognizing Likurai Dance movements remains challenging, particularly in detecting the specific dance movements. This research aims to test the effectiveness of detecting three types of Likurai Dance movements using documented digital video. The detection model is the YOLOv8 algorithm, known for detecting objects quickly and accurately. A YOLOv8-based platform is proposed to detect these dance movements precisely. In the testing, the YOLOv8 model demonstrated outstanding performance, achieving a very high mAP of 99.5% for the Wesei Wehali movement, 99.4% for the Be Tae Be Tae movement, and 99.1% for the Tebe Re movement. These results indicate that the model can detect dance movements with exceptional accuracy, precision, and recall rates above 98%. This research concludes that YOLOv8 has excellent potential in detecting traditional dance movements with high accuracy. These findings are significant for preserving and documenting the Likurai Dance and provide an educational means for younger generations to understand better and appreciate traditional cultural values.
Improved Hybrid GoogLeNet-Based Deep Learning Optimization for Standardized Straw Mushroom Quality Classification in Indonesia Priyatna, Bayu; Abdurahman, Titik Khawa; Miskon, Muhammad Fahmi; Hananto, April Lia; Hananto, Agustia Tia; Rahman, Aviv Yuniar
Journal of Applied Data Sciences Vol 7, No 2: May 2026
Publisher : Bright Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47738/jads.v7i2.1206

Abstract

Deep learning plays a crucial role in modern computer vision due to its ability to automatically extract hierarchical features from large-scale image data. Among various architectures, Convolutional Neural Networks (CNNs) have been extensively utilized for image pattern interpretation, including in agricultural product inspection. Straw mushrooms (Volvariella volvacea) are important agro-industrial commodities in Indonesia; however, their quality assessment still relies on subjective manual evaluation based on the Indonesian National Standard (SNI:01-6945-2003), leading to inconsistency in grading results. To address this limitation, this research proposes an Improved Hybrid GoogLeNet model integrated with a YOLO-based detection framework and hybrid preprocessing to enhance feature clarity and classification robustness. The system is capable of conducting object detection, 3-class morphological quality classification (Pure White, Oval, and Black Spot/Defect), and automatic diameter measurement using calibrated pixel-to-centimeter conversion. Performance evaluation is carried out by benchmarking the proposed model against several popular deep learning architectures including YOLOv5, LeNet, AlexNet, VGGNet, and ResNet. Experimental results demonstrate that the Improved Hybrid GoogLeNet achieves the highest performance with precision of 97.99%, recall of 96.07%, and F1-score of 96.98%, along with low misclassification rates across all classes. These results indicate that the proposed method provides accurate, reliable, and efficient quality assessment that supports standardized automated grading in industrial applications. Therefore, this study contributes to the advancement of intelligent computer vision solutions for digital transformation in the Indonesian mushroom agro-industry.