Article Per Year (5 Year)
p-Index From 2021 - 2026
0.444
P-Index
This Author published in this journals
All Journal Jurnal Mantik Jurnal Sistem Informasi Triguna Dharma (JURSI TGD)
Nurhayati
Universitas Potensi Utama
Computer Science & IT Economics, Econometrics & Finance Languange, Linguistic, Communication & Media

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 2 Documents
Search

 1 
Bipolar Approach in Recognition of Gorga Batak Patterns with the Hebbian Method Frinto Tambunan; Eviyanti Novita Purba; Nurhayati; Juliana Naftali Sitompul
Jurnal Mantik Vol. 5 No. 4 (2022): February: Manajemen, Teknologi Informatika dan Komunikasi (Mantik)
Publisher : Institute of Computer Science (IOCS)

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

Abstract

The recognition of shapes or patterns in Intelligent Artificial is a method that is developing very rapidly and is continuously being developed to this day. The need for information contained in a pattern or form is very useful to be developed and applied properly to the health or culture sector, by using a system that is embedded with a capability or feature in pattern recognition. This model can be applied both to face pattern recognition, patterns. Likewise in this discussion about the pattern recognition of the Gorga Batak using artificial neural networks, namely the Hebbian method. With two patterns as knowledge or learning base and then tested. The input pattern will be checked for the similarity of the two learning base patterns, whether it is recognized as the Gorga “Simeol – eol” pattern or as the Gorga “Sitompi” pattern. By using 25 input variables and bias 1 with an initial weight value of 0, the Gorga “Someol-eol” and “Sitompi” patterns were initialized to a grayscale image and then extracted to a bipolar image with values of 1 and -1. The Gorga “Simeol – eol” pattern has a target of 1 and the Gorga “Sitompi” pattern has a target of -1, the function f (net) is 1 if Y > = 0 and -1 if Y < 0. From the process carried out, it is found that the Gorga “Simeol – eol” pattern is obtained the value of Y = 12 and for the Gorga “Sitompi” pattern Y = -12, and inputted to the function f (net), the result is the same as the target 1 for the Gorga “Simeol –eol” pattern and -1 for the Gorga “Sitompi” pattern.
Analisis Perbandingan Optimasi Seleksi Fitur Logistic Regression dan SVM untuk Prediksi PCOS Annisa Ashari; Lumi Krismona; Nurhayati
Jurnal Sistem Informasi Triguna Dharma (JURSI TGD) Vol. 5 No. 2 (2026): EDISI MARET 2026
Publisher : STMIK Triguna Dharma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.53513/jursi.v5i2.12476

Abstract

Polycystic Ovary Syndrome (PCOS) merupakan gangguan endokrin pada wanita usia reproduktif yang ditandai ketidakteraturan menstruasi, hiperandrogenisme dan perubahan morfologi ovarium, serta berpotensi menimbulkan infertilitas dan komplikasi metabolik (WHO, 2025). Penelitian ini bertujuan menganalisis secara komparatif kinerja algoritma Logistic Regression dan Support Vector Machine (SVM) yang dioptimasi dengan feature selection untuk prediksi PCOS berbasis data klinis. Dataset berisi 1.000 data pasien dengan lima atribut klinis, yaitu umur, indeks massa tubuh (BMI), ketidakteraturan menstruasi, kadar testosteron dan jumlah folikel antral, serta label biner diagnosis PCOS. Data dibagi menggunakan stratified train-test split 80:20 dan seluruh fitur numerik dinormalisasi. Optimasi dilakukan dengan mengintegrasikan Recursive Feature Elimination (RFE) dan Grid Search pada Logistic Regression untuk menentukan kombinasi jumlah fitur dan parameter regulasi terbaik, sementara pada SVM dilakukan penalaan parameter C, jenis kernel dan gamma menggunakan GridSearchCV dengan 5-fold stratified cross-validation dan F1-score sebagai metrik acuan, mengikuti praktik optimasi model yang banyak digunakan pada penelitian PCOS berbasis machine learning. Hasil eksperimen menunjukkan bahwa Logistic Regression terbaik mencapai akurasi 0,915, F1-score kelas PCOS 0,80 dan AUC 0,978, sedangkan SVM memberikan kinerja lebih tinggi dengan akurasi 0,97, F1-score kelas PCOS 0,92 dan AUC 0,998. Secara keseluruhan, hasil ini mengindikasikan bahwa SVM dengan fitur terpilih lebih efektif dibanding Logistic Regression, selaras dengan beberapa studi yang melaporkan keunggulan model SVM dalam deteksi PCOS.
Co-Authors Annisa Ashari Frinto Tambunan Lumi Krismona Purba, Eviyanti Novita