Analisis Prediksi Keterlambatan Pembayaran Listrik Menggunakan Komparasi Metode Klasifikasi Decision Tree dan Support Vector Machine

Abstract

Electrical energy is one of the most needed energy today. In this modern era, almost all human activities cannot be separated from the use of electricity. The only electricity supply company in Indonesia is the State Electricity Company or PT. PLN (Persero). PLN also has several obstacles. One of them is the very large amount of arrears to customers. This causes considerable losses for PLN. The electricity payment counter also experienced the same thing as experienced by PLN, as at PT Jaya Nuhgra Pratama. For this reason, this research was conducted by comparing two methods in the classification, namely decision tree and support vector machine to find out which method is the best in solving problems and to find out the factors that are the main causes of delays in electricity payments

The Shopee Application User Reviews Sentiment Analysis Employing Naïve Bayes Algorithm

“…Furthermore, diverse comparisons between classification methods have been conducted. Noteworthy studies include the improvement of sales performance with decision tree achieving 100% accuracy compared to Naïve Bayes with 41% [10], classification of fish freshness using K-Nearest Neighbor and Hue Saturation Value achieving 93% accuracy [11], and Sentiment analysis on Shopee user reviews, where Naïve Bayes displayed high accuracy at 99.5% [12]. Comparisons between C4.5 and Naïve Bayes in predicting student graduation showed distinct performances, with C4.5 having 76.69% accuracy and Naïve Bayes with 72.95% [13].…”

“…Research has revealed diverse applications and comparisons between different classification methods in various contexts. One study focused on improving sales performance with decision tree and naïve Bayes methods, finding that decision tree achieved 100% accuracy while naïve Bayes only 41% [10]. On the other hand, other research such as fish freshness classification using K-Nearest Neighbor and Hue Saturation Value managed to achieve 93% accuracy [11], while in Sentiment analysis of user reviews on Shopee, the Naïve Bayes algorithm gave high accuracy results of 99.5 % [12].…”

Perbandingan Metode K-NN, Naïve Bayes, Decision Tree untuk Analisis Sentimen Tweet Twitter Terkait Opini Terhadap PT PAL Indonesia

“…Metode ini juga merupakan salah satu metode data mining dan termasuk dalam metode machine learning yang memiliki akurasi yang tinggi dan merupakan salah satu algoritma yang sederhana [15]. Tahapan metode Naïve Bayes sebagai berikut [16] : a) Menghitung nilai probabilitas bersyarat: Decision Tree merupakan salah satu metode data mining untuk mengklasifikasikan data yang terdiri dari root node, internal nood dan terminal nood [7] [17]. Variabel atau fitur adalah root node dan internal nood sedangkan label kelas adalah terminal nood [7].…”