Claim Missing Document
Check
Articles

Found 3 Documents
Search

Perbandingan Akurasi Metode Naive Bayes dan Metode Random Forest dalam Mendiagnostik Penyakit Kanker Payudara Rohmah, Nisaur; Safitri, Eka Ayu; Alinta, Cici; Oktalina, Yuyun; Setiawan, Wahyudi
DoubleClick: Journal of Computer and Information Technology Vol. 8 No. 2 (2025): Edisi Februari 2025
Publisher : Universitas PGRI Madiun

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25273/doubleclick.v8i2.20383

Abstract

Kanker payudara merupakan penyakit yang mempunyai dampak signifikan terhadap kesehatan wanita di seluruh dunia. Dalam penelitian ini, kami menggunakan data dari kumpulan data Wisconsin Diagnostic untuk membandingkan kinerja dua metode klasifikasi dalam diagnosis kanker payudara: Naive Bayes dan Random Forest. Metode Naive Bayes menggunakan pendekatan probabilitas sederhana untuk mengklasifikasikan data, sedangkan Random Forest membuat beberapa pohon keputusan dan menggabungkannya untuk meningkatkan akurasi prediksi. Sebelum membandingkan performa kedua algoritma, dilakukan proses preprocessing data yang meliputi identifikasi fitur, deteksi outlier, dan normalisasi. Kedua algoritma tersebut dilatih dan diuji menggunakan data dari dataset Wisconsin Diagnostic yang terdiri dari 569 sampel dengan 32 atribut dan 2 kelas. Hasil pengujian menunjukkan bahwa Naive Bayes menghasilkan akurasi sebesar 0.932, precision sebesar 0.933, recall sebesar 0.931, dan F1-score sebesar 0.932. Sedangkan Random Forest menghasilkan akurasi sebesar 0.9442, precision sebesar 0.945, recall sebesar 0.944, dan F1-score sebesar 0.944. Dalam konteks ini, Random Forest sedikit lebih akurat dibandingkan Naive Bayes, namun keduanya baik untuk mendiagnosis kanker payudara jinak dan ganas. Penelitian ini memberikan wawasan tentang efektivitas Naive Bayes dan Random Forest dalam membantu diagnosis kanker payudara berdasarkan kumpulan data klinis modern. Meskipun hasilnya menunjukkan bahwa Random Forest lebih unggul dalam hal akurasi, penting untuk mempertimbangkan konteks dan karakteristik kumpulan data ketika memilih algoritma klasifikasi yang tepat untuk aplikasi medis.
The Influence of Brand Awareness, Brand Image Recognition on Brand Loyalty with the Intervening Mediation of Brand Trust and Brand Love on Samsung Smartphone Products in Indonesia Safitri, Eka Ayu; Albari
Indonesian Journal of Economics, Business, Accounting, and Management (IJEBAM) Vol 2 No 4 (2024): Volume 2, No. 4, 2024
Publisher : PT SOLUSI EDUKASI BERDIKARI

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.63901/ijebam.v2i4.67

Abstract

This research explains the influence of brand awareness, brand image, brand trust, brand love on brand loyalty of Samsung smartphones in Indonesia. Globalization has brought about changes in the world of technology, one of which is the Samsung folding smartphone model. This research used 195 samples. Sampling used a non-probability method with convenience sampling techniques. The analysis techniques used are structural equation modeling (SEM) and AMOS. The research results concluded that there is an influence of: brand awareness on brand image; brand image impacts brand trust; brand trust influences brand love; brand trust impacts brand loyalty; brand love influences brand loyalty. This research also shows that there is no influence of: brand awareness on brand trust; brand awareness has no effect on brand love; brand image has no impact on brand love; brand awareness has no impact on brand loyalty; and brand image has no impact on brand loyalty.
BATASAN DAN MEKANISME PENERAPAN SANKSI PIDANA PERPAJAKAN DI INDONESIA DALAM PERSPEKTIF ASAS ULTIMUM REMEDIUM Safitri, Eka Ayu; Damayanti, Ratih; Sulistiyono, Tri
Jurnal Hukum Statuta Vol 4 No 3 (2025): Volume 4, Nomor 3, Agustus 2025
Publisher : Fakultas Hukum Universitas Pembangunan Nasional Veteran Jakarta

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35586/jhs.v4i3.11160

Abstract

This study examines the limitations and mechanisms of the application of criminal sanctions against taxpayers who violate tax provisions in Indonesia, and evaluates the application of the ultimum remedium principle as the principle of "criminal sanctions as a last resort." Using a qualitative legal-normative approach with a descriptive-analytical method, data was collected through a documentation study of primary sources (Law 7/2021 in conjunction with Law 6/2023 concerning KUP, KUHAP, KUHP) and secondary sources (scientific journals 2020–2025, textbooks, DGT guidelines). Normative-dogmatic analysis reveals that the KUP Law explicitly separates administrative and criminal sanctions without setting a minimum loss threshold for criminal prosecution, so that tax officials tend to use criminal sanctions only for high-value cases. The legal-theoretical study explains that the ultimum remedium principle is accommodated through provisions such as Article 8 paragraph (3) and Article 44B of the KUP Law, but in practice it still encounters inconsistencies and broad discretion. Synthesis of findings indicates the need for quantitative guidelines for loss thresholds, reintegration of the “first-time offender” protocol, and harmonization of the Criminal Procedure Code with the principle of criminal subsidiarity. These recommendations are expected to strengthen the implementation of the ultimum remedium principle, ensuring that criminal sanctions are truly the last resort in enforcing fair and proportional tax law. Keywords: Taxation; Ultimum Remedium; Criminal Sanctions; KUP Law; Criminal Law