Article Per Year (5 Year)
p-Index From 2021 - 2026
0.751
P-Index
This Author published in this journals
All Journal METIK JURNAL Jurnal Mantik Progresif: Jurnal Ilmiah Komputer
Deddy Kurniawan
Universitas Mulia
Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Earth & Planetary Sciences Economics, Econometrics & Finance Electrical & Electronics Engineering Languange, Linguistic, Communication & Media

Published : 4 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 4 Documents
Search

 1 
Implementasi Association Rule Dalam Menganalisis Data Penjualan Sheshop dengan Menggunakan Algoritma Apriori Sherina Aulia Miranda; Fahrullah Fahrullah; Deddy Kurniawan
METIK JURNAL Vol 6 No 1 (2022): METIK Jurnal
Publisher : LPPM Universitas Mulia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47002/metik.v6i1.342

Abstract

Sheshop is a business activity engaged in the field of hamper making services. Since the last 2 (two) years, sales transactions in Sheshop have been increasing, the transaction data is only used as a report and is not used to regulate business strategies. The transaction data should be used to see the attachment of each type of product purchased by the customer simultaneously. The amount of sales transaction data on Sheshop can be used as an analysis of customer behavior in making purchases of hampers at Sheshop. This study performs data analysis by implementing the Apriori algorithm method because this algorithm handles data mining processes quickly on large amounts of data, from the results of this study Sheshop can make decisions on what items need more inventory compared to other items by looking at the value confidence and support by using the RapidMiner application. The results of this study indicate that the association rule formed from 568 Sheshop sales data uses a minimum support value of 10% and a minimum confidence of 50% produces 6 (six) association rules with a confidence value of 58% to 75% with all rules having a positive correlation level. Based on the 6 (six) association rules obtained, 2 products are often purchased at the same time, namely the Koran and tasbih with a confidence value of 75%.
Selection Participants Of Science Olympic In Elementary School Using Fuzzy – Profile Matching Method I Komang Wiratama; Tuti Marjan Fuadi; Emy Yunita Rahma Pratiwi; Deddy Kurniawan; I Gede Iwan Sudipa
Jurnal Mantik Vol. 6 No. 2 (2022): August: Manajemen, Teknologi Informatika dan Komunikasi (Mantik)
Publisher : Institute of Computer Science (IOCS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35335/mantik.v6i2.2671

Abstract

Science ability is one of the students' potentials that can be developed and can be competed. Every student's potential in the field of science can be excelled and directed to become ability. Science Olympiad is a forum to show students' abilities so that every elementary school is expected to participate. Schools must, of course, determine prospective participants for the science olympiad, with the selection of potential students with predetermined assessment criteria. In this study, implementing decision support with fuzzy logic for the conversion of criteria values ??and the Profile Matching method in determining the ideal profile of prospective Olympic participants, each student's alternative value on the criteria is obtained by the difference in value with the ideal profile, the smaller the value of the GAP difference, it indicates the best alternative. The results showed that the results of fuzzy and profile decisions were the three best alternatives for students who participated in the science Olympics.
Optimalisasi Strategi Pemenuhan Persediaan Stok Barang Menggunakan Algoritma Frequent Pattern Growth Deddy Kurniawan; Maurits Sahata Sipayung; Rika Ismayanti; Muhammad Rivani Ibrahim; Yeva Bintan; Sherina Aulia Miranda
METIK JURNAL Vol 6 No 2 (2022): METIK Jurnal
Publisher : LPPM Universitas Mulia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47002/metik.v6i2.387

Abstract

Pemenuhan kebutuhan stok persediaan barang merupakan salah satu dari pilar utama proses bisnis yang rutin dilakukan pelaku bisnis secara umum. Peluang akan terjadinya kesalahan perhitungan yang dilakukan secara konvensional tanpa adanya sebuah analisis mendalam yang menyebabkan tidak akuratnya penentuan jumlah persediaan yang harus dipenuhi. Hasil penelitian menyajikan sebuah solusi dengan pendekatan Data Mining menggunakan teknik aturan asosiasi (association rule). Pendekatan data mining dibangun dengan menggunakan sebuah kerangka kerja pupuler data mining CRoss Industry Standard Process for Data Mining (CRISP-DM) yang dikerjakan dalam 6 tahapan yaitu Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation, dan Deployment. Sampel UMKM kota samarinda menjadi objek pada penelitian dengan menggunakan 1000 data dari riwayat transaksi penjualan dalam kurun waktu tertentu yang diidentifikasi dengan menjalankan algoritma Frequent Pattern Growth (FP-Growth) untuk memaksimalkan kinerja komputasi dalam proses ekstraksi pola item barang. Ekstraksi pola aturan dari dataset transaksi penjualan dilakukan dengan 9 kali percobaan dengan melakukan perubahan nilai support (S) dan confidence (C) dengan hasil percobaan trbaik menghasilkan 9 best rule dengan rentang nilai S sebesar 9% - 14% dan C sebesar 60% - 75% yang mencakup aturan 2-itemset dan 3-itemset. Masing-masing rule diterapkan uji lift yang menghasilkan rentang nilai 2.790 – 3.698 dengan rata-rata nilai lift sebesar 3.26, dimana setiap aturan memenuhi nilai minimum (lift > 1.00) yang menunjukkan setiap kombinasi aturan memiliki peluang cross-selling yang baik
Meningkatkan Efisiensi Prediksi Risiko Diabetes Melitus dengan Metode Fuzzy Decision Tree Deddy Kurniawan; Wahyu Yanuarta; Tina Tri Wualansari; Niken Ayu Dwi Febrianti
Progresif: Jurnal Ilmiah Komputer Vol 21, No 2 (2025): Agustus
Publisher : STMIK Banjarbaru

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35889/progresif.v21i2.2680

Abstract

The main problem in Fuzzy Logic (FL)-based prediction models is that the number of rules increases as the data dimension increases, reducing the efficiency of the system in interpretation and prediction. This study aims to unravel the complexity and improve the accuracy value of DM predictions using the Fuzzy Decision Tree (FDT) method based on Iterative Dichotomiser 3 (ID3). The research data was obtained from the National Institute of Diabetes and Digestive and Kidney Diseases with the parameters of glucose, BMI, HDL, and systolic blood pressure. The process includes data fuzzification, the formation of a decision tree with ID3, and the application of two thresholds, namely FCT and LDT. The results showed that the FDT model succeeded in reducing the number of rules by 25%, from 81 rules to 60 rules. The application of ID3-based FDT succeeded in increasing the accuracy value of DM predictions by 80%. The conclusion of the study states that the FDT model is able to unravel the complexity of the prediction model by using a simpler number of rules and can maintain and increase the accuracy value of the DM prediction model.Keywords: Diabetes Melitus; Fuzzy Decision Tree; Fuzzy Logic; Fuzzy Sugeno; Risk Prediction AbstrakPermasalahan utama dalam model prediksi berbasis Fuzzy Logic (FL) adalah meningkatnya jumlah aturan seiring bertambahnya dimensi data, mengurangi efisiensi sistem dalam interpretasi dan prediksi. Penelitian ini bertujuan mengurai kompleksitas dan meningkatkan nilai akurasi prediksi DM menggunakan metode Fuzzy Decision Tree (FDT) berbasis Iterative Dichotomiser 3 (ID3). Data penelitian diperoleh dari National Institute of Diabetes and Digestive and Kidney Diseases dengan parameter glukosa, BMI, HDL, dan tekanan darah sistolik. Proses meliputi fuzzifikasi data, pembentukan pohon keputusan dengan ID3, serta penerapan dua threshold, yaitu FCT dan LDT. Hasil penelitian menunjukkan bahwa model FDT berhasil mengurangi jumlah aturan sebesar 25%, dari 81 aturan menjadi 60 aturan. Penerapan FDT berbasis ID3 berhasil meningkatkan nilai akurasi prediksi DM sebesar 80%. Simpulan penelitian menyatakan bahwa model FDT mampu untuk mengurai kompleksitas model prediksi dengan menggunakan menghasilkan jumlah aturan yang lebih sederhana dan dapat menjaga serta meningkatkan nilai akurasi model prediksi DM.Kata kunci: Diabetes Melitus; Pohon Keputusan Fuzzy; Logika Fuzzy; Fuzzy Sugeno; Prediksi Risiko
Co-Authors Emy Yunita Rahma Pratiwi, Emy Yunita Rahma Fahrullah, Fahrullah I Gede Iwan Sudipa I Komang Wiratama Ibrahim, Muhammad Rivani Maurit Sipayung Niken Ayu Dwi Febrianti Rika Ismayanti Sherina Aulia Miranda Tina Tri Wualansari Tuti Marjan Fuadi Wahyu Yanuarta Yeva Bintan