Garuda - Garba Rujukan Digital

COMBINING BOOTSTRAPPING AND GENETIC ALGORITHM BASED ON FEATURE SELECTION FOR FRANCHISE LOCATION PROSPECT PREDICTION Tati Mardiana
Jurnal Riset Informatika Vol 3 No 3 (2021): Period of June 2021
Publisher : Kresnamedia Publisher

Location selection is crucial in the franchise fast-food industry. A thorough location selection model paired with a proper analytical technique can considerably improve the performance of placement decisions, attract more customers, and raise market share and profitability. Franchise location data sets have an imbalanced class nature. The franchise location prospect prediction performance decreased as a result of the dataset's noisy characteristics. We developed a hybrid approach to improve franchise location prospect prediction performance in this study. It combines Bootstrapping to address class imbalance problems and Genetic Algorithm (GA) to select relevant features in the franchise location prospect prediction. We experimented with four different classification methods (Naive Bayes, C4.5, Random Forest, ID3, Gradient Boosted Trees). The results show that almost all classifiers that use Bootstrapping and GA outperform the original technique. We employ the Confusion Matrix and Root Mean Squared Error (RMSE) to examine the proposed method's performance. The test results demonstrate that the proposed method considerably enhances the franchise location prospect's classification performance.

OPTIMASI NAÏVE BAYES DENGAN PARTICLE SWARM OPTIMIZATION DAN STRATIFIED UNTUK PREDIKSI KREDIT MACET PADA KOPERASI Tati Mardiana
Jurnal Riset Informatika Vol. 1 No. 1 (2018): Periode Desember 2018
Publisher : Kresnamedia Publisher

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

Dalam bisnis, koperasi memiliki peranan penting dalam meningkatkan perekonomian nasional. Ketidakmampuan anggota untuk membayar angsuran kredit merupakan masalah utama yang terjadi pada koperasi. Akibatnya, terjadi kredit macet. Koperasi dapat menghindari kredit macet dengan membuat prediksi dari anggota koperasi yang berpotensi terlambat membayar kredit. Dalam beberapa penelitian telah menggunakan Naive Bayes untuk masalah klasifikasi karena perhitungan yang efisien, dan akurasi tinggi. Tetapi Naive Bayes mengasumsikan bahwa semua atribut kelas tidak tergantung pada atribut lainnya. Naive Bayes sesuai untuk masalah klasifikasi dengan atribut besar. Namun, asumsi ini sering tidak dapat dipertahankan dalam masalah klasifikasi nyata. Dalam beberapa dokumen, kinerja Naive Bayes tidak sempurna. Tujuan dari penelitian ini adalah untuk mengoptimalkan metode Naive Bayes menggunakan Particle Swarm Optimization (PSO) dan untuk meningkatkan akurasi dalam memprediksi kredit macet di koperasi. Penelitian ini menggunakan data dari Pusat Data Koperasi (PUSKOPDIT) DKI Jakarta. Data set kredit yang diperoleh sebanyak 565 record dengan 15 prediktor atribut dan 1 atribut kelas. Hasil pengujian dengan confusion matrix dan kurva ROC diperoleh dari nilai akurasi sebesar 86% dan nilai sebesar 0,867 dengan diagnosis klasifikasi baik. Penelitian ini menunjukkan bahwa penggunaan PSO pada NBC untuk memprediksi kredit macet meningkatkan akurasi 21,03% dan AUC sebesar 0,069. Hasil uji T-Test dan Anova menunjukkan bahwa pada dua metode klasifikasi yang diuji memiliki perbedaan yang nyata (signifikan) dalam nilai AUC.

SISTEM PENDUKUNG KEPUTUSAN PEMILIHAN PERGURUAN TINGGI SWASTA MENGGUNAKAN TOPSIS Tati Mardiana; Siska Sivia Tanjung
Jurnal Riset Informatika Vol. 1 No. 2 (2019): Periode Maret 2019
Publisher : Kresnamedia Publisher

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

Memilih perguruan tinggi yang tepat merupakan langkah penting bagi siswa dalam mempersiapkan karir dan masa depannya. Dengan pendidikan di perguruan tinggi, siswa meningkatkan kesempatannya untuk mendapatkan pekerjaan yang lebih baik. Tetapi keterbatasan daya tampung perguruan tinggi negeri membuat siswa dan orang tua harus memilih perguruan tinggi swasta yang sesuai dengan keinginan dan kemampuan. Kesalahan dalam memilih perguruan tinggi menyebabkan siswa mengalami kegagalan dalam menjalankan pendidikan di perguruan tinggi tersebut. Oleh karena itu, siswa dan orang tua perlu mempertimbangkan beberapa faktor seperti status akreditasi, biaya, jumlah mahasiswa, jumlah dosen, fasilitas, program studi dan lain-lain untuk memilih perguruan tinggi swasta. Kendati demikian, banyak siswa dan orang tua yang mengalami kebingungan dalam memilih perguruan tinggi swasta. Hal ini disebabkan karena banyaknya perguruan tinggi swasta dan minimnya informasi tentang perguruan tinggi swasta tersebut. Tujuan penelitian ini adalah membangun sistem pendukung keputusan pemilihan perguruan tinggi swasta yang sesuai keinginan dan kemampuan siswa dan orang tua. Penelitian ini menggunakan logika Fuzzy Multiple Attribute Decision Making (FMADM) dengan metode Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) untuk melakukan perangkingan dari setiap alternatif perguruan tinggi swasta. Hasil pengujian menunjukkan bahwa kinerja sistem memenuhi persyaratan fungsional dan kinerja sistem mencapai akurasi sebesar 83,33%. Sistem pendukung keputusan ini membantu siswa dan orang tua membuat keputusan untuk memilih perguruan tinggi swasta yang sesuai dengan keinginan dan kemampuan mereka secara akurat.

DECISION SUPPORT SYSTEM FOR DETERMINING APPROPRIATE FRANCHISE LOCATIONS USING THE PROFILE MATCHING METHOD Tati Mardiana; Yesni Malau
Jurnal Riset Informatika Vol. 3 No. 1 (2020): December 2020 Edition
Publisher : Kresnamedia Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34288/jri.v3i1.48

Finding the appropriate location is crucial when starting a franchise. The appropriate location will affect the overall business risk and profitability of the franchise. Nevertheless, some franchises have a bankruptcy in running their business. One of the factors that contribute to the bankruptcy of a franchise business is a location that does not meet several criteria that support business success. Therefore, this study aims to propose a decision support system model to determine the location of the franchise based on matching profiles between the actual data value of a location and the value of the location profile expected by the franchisor. The profile matching method has a better level of objectivity because it measures the value of each indicator variable. The criteria for determining the location of a franchise are potential customers, access to location, competition, and costs. The test results show that the decision support system to determine the location of the franchise using the profile matching method meets the functional requirements. This decision support system helps franchises to determine the appropriate when starting a franchise.

An Expert System for Detection of Diabetes Mellitus with the Forward Chaining Method Tati Mardiana; Ega Maulana Ditama; Tuslaela Tuslaela
Jurnal Riset Informatika Vol. 2 No. 2 (2020): March 2020 Edition
Publisher : Kresnamedia Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34288/jri.v2i2.49

In recent years, diabetes mellitus in Indonesia has become a health problem in the community because its population has increased 2-3 times faster than in other countries. Diabetes prevalence in Indonesia ranks 4th highest globally after China, India, and the United States. People can prevent complications and premature death if they detect early symptoms of diabetes. However, people do not know that they are at risk of diabetes and do not have knowledge about the symptoms of diabetes, the complexity of the process of diagnosis, and the high cost of examinations. Therefore, we need an application that can provide the results of the type of diabetes and its management solutions as practiced by experts. This research aims to develop an expert system for detecting types of diabetes such as type one diabetes, type two diabetes, neuropathy diabetes, diabetes retinopathy, and diabetes nephropathy. The object of this research is diabetes, which was carried out from March to April 2019 in the Klinik Pratama Desa Putera. This study uses primary data from patients with a history of diabetes at Klinik Pratama Desa Putra and secondary data from literature, research journals, and data documents needed to compile this study. In addition, we generated a knowledge base using forward chaining. The test results show that the expert system meets the functional requirements, and the system performance reaches an accuracy of 100%. This expert system helps people in Indonesia to detect diabetes early so that it can prevent complications.

COMBINING BOOTSTRAPPING AND GENETIC ALGORITHM BASED ON FEATURE SELECTION FOR FRANCHISE LOCATION PROSPECT PREDICTION Tati Mardiana
Jurnal Riset Informatika Vol. 3 No. 3 (2021): June 2021 Edition
Publisher : Kresnamedia Publisher

Location selection is crucial in the franchise fast-food industry. A thorough location selection model paired with a proper analytical technique can considerably improve the performance of placement decisions, attract more customers, and boost market share and profitability. Franchise location data sets have an imbalanced class nature. The franchise location prospect prediction performance decreased as a result of the dataset's noisy characteristics. We developed a hybrid approach to improve franchise location prospect prediction performance in this study. It combines Bootstrapping to address class imbalance problems and a Genetic Algorithm (GA) to select relevant features in the franchise location prospect prediction. We experimented with five different classification methods (Naive Bayes, C4.5, Random Forest, ID3, Gradient Boosted Trees). The results show that almost all classifiers that use Bootstrapping and GA outperform the original technique. We employ the Confusion Matrix and Root Mean Squared Error (RMSE) to examine the proposed method's performance. The test results demonstrate that the proposed method considerably enhances the franchise location prospect's classification performance.

Sentiment Analysis of Digital Wallet Service Users Using Naïve Bayes Classifier and Particle Swarm Optimization Alvie Delia Cahyani; Tati Mardiana; Laela Kurniawati
Jurnal Riset Informatika Vol. 2 No. 4 (2020): Period September 2020
Publisher : Kresnamedia Publisher

Digital wallet services adequately provide many benefits to their users. However, not all users of digital wallet services have a favourable opinion of the service. Therefore, online transportation service companies need to carry out an analysis to determine general sentiment towards their products. The Naïve Bayes Classifier method represents a simple, fast method with excellent accuracy and performs comparatively well for classifying data. However, the Naïve Bayes Classifier method assumes that the attributes are independent, so it can cause the accuracy to be less than optimal. This study aims to improve the accuracy of the Naive Bayes classification for classifying public opinion on digital wallet services using Particle Swarm Optimization. This study manages data from Twitter as much as 490 tweet data. The test results with the confusion matrix and ROC curves show an increase in the accuracy of the Naïve Bayes Classifier method for the Dana digital wallet from 60.00% to 91.67% and the iSaku digital wallet from 53.23% to 85.00%. The T-test and ANOVA test results show that the test results of both classification methods provide significant differences in the accuracy value.

Title

Found 7 Documents
Search
Journal : Jurnal Riset Informatika

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title Search

Found 7 Documents Search Journal : Jurnal Riset Informatika

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Abstract

Title

Found 7 Documents
Search
Journal : Jurnal Riset Informatika