Penerapan Algoritma C4.5 Untuk Prediksi Loyalitas Nasabah PT Erdika Elit Jakarta

Abstract

C4.5 algorithm is a decision tree algorithm group. This algorithm has input in the form of training samples and samples. While samples are data fields which we will use as parameters in classifying data. From the variable transaction frequency the company can see which customers are loyal to the company based on historical customer transaction data, but there are still some variables that make customers loyal to the company. These variables are age, customer gender, company sales gender, educational background, customer transaction frequency. The company knows how to predict customers who will be loyal to the company based on the experience of some of the variables above, but the company does not know the most influential variable in the assessment of loyal customers because of some of the variables above are not interconnected and it is uncertain if one variable can make a decision whether the customer loyal. Based on the decision tree that has made the most influential attribute on customer loyalty is the educational background because it has the highest gain value of 1.545292721 and as the root of the decision tree while the client's gender does not significantly affect customer loyalty because it is always at the last node with the gain value which is 0.623919119.

Implementation of Data Mining Using C4.5 Algorithm for Predicting Customer Loyalty of PT. Pegadaian (Persero) Pati Area Office

“…The author uses the C 4.5 algorithm on the basis of some literature from previous research from Teguh Budi Santoso with the title "Analisa Dan Penerapan Metode C4.5 Untuk Prediksi Loyalitas Pelanggan" with the results showing that classification with the C4.5 algorithm gets an accuracy of up to 97.5%, which means that the exact C4.5 algorithm is used to calculate the level of customer loyalty [1]. The second study entitled "Penerapan Algoritma C4.5 Untuk Prediksi Loyalitas Nasabah PT Erdika Elit Jakarta" by Khotibul Umam, based on the decision tree that has been made the attribute that has the most influence on customer loyalty is educational background because it has the highest gain value, namely 1.545292721 and as the root of the decision tree while the gender of the customer does not have much effect on customer loyalty because it is always at the last node with a gain value of 0.623919119 [2]. The third study entitled "Penerapan algortima c4.5 untuk penentuan kelayakan kredit" by Siti Nur Khasanah, Based on the results of the application of the C4.5 classifier algorithm, it can be concluded that to determine credit worthiness whether a prospective customer will become a customer with smooth or problematic payments using the C4.5 algorithm classifier and the accuracy of the C4.…”

Optimalisasi Algoritma C4.5 dalam Menganalisis Indikasi Penyebab Penyakit Feline Immunodeficiency Virus (FIV) pada Kucing

“…Algoritma C4.5 memiliki ide dasar untuk membentuk pohon keputusan. Pohon keputusan termasuk dalam klasifikasi dan prediksi terkenal karena relatif mudah dipahami dengan bahasa alami sehingga dapat diinterpretasikan dengan cepat [17].…”

Clustering and Modeling of E-Government Architecture Using the Algorithm C4.5

“…Algoritma ini mempunyai input berupa training samples dan samples. Sedangkan samples merupakan field-field data yang nantinya akan kita gunakan sebagai parameter dalam melakukan klasifikasi data [5]. Salah satu algoritma yang dipakai dalam data mining adalah algoritma C4.5 [6].…”