Komparasi Metode Decision Tree, KNN, dan SVM Untuk Menentukan Jurusan Di SMK

Abstract

The selection of majors for prospective vocational students is the first step in determining the next career. The determination of the major aims so that students can be directed in receiving lessons based on the ability and talent of students and of course when they have graduated have the skills to get a job if they do not continue their studies. Siti Banun Sigambal Private Vocational School is located in Labuhanbatu Rantau Prapat. In realizing one of the missions of SMK, namely Realizing quality learning in Vocational High School, SMK Siti Banun in determining the Department by conducting tests. In classifying data mining techniques can be used, among others Decision Tree, K-Nearest Neighbors (KNN) and Support Vector Machine (SVM). This research was conducted to compare the performance of Decision Tree, KNN and SVM algorithms in determining majors. Of the 245-test data used obtained SVM has an accuracy value of 89%, precision 87% while KNN has an accuracy value of 84%, precision 81% and Decision Tree has an accuracy value of 78% and precision of 75%.

Sentiment Analysis on Hotel Ratings Using Dynamic Convolution Neural Network

“…This algorithm has the advantage of high accuracy and does not require a lot of data samples to avoid overfitting [7]. This algorithm can also solve problems by using datasets that have a large feature space [8]. In addition to SVM, there is also another algorithm that is also good for use in prediction problems in the form of classification, namely the perceptron, a simple supervised learning neural network algorithm that can be used to recognize patterns in data [9].…”

The Factors that Affect Customer Ratings in E-Commerce Company

“…This algorithm uses a K value to measure the number of instances with similar values, with a distance metric method as the measuring tool [24]. Some studies have shown the performance of this algorithm in solving classification problems, such as the classification of vocational school's major, with the values: accuracy = 84%, precision = 81%, and recall = 84% [25]; classifying the electrical subsidies' recipients for the household, with an accuracy value of 98.07% [26]; and the the Iris flower classification using the K value of 5, with an accuracy value of 96.677% [27]. These studies are the reference we use in this research, both in the classification steps and choosing the K value.…”

Classification of Shape Bean Coffee Using Convolutional Neural Network

“…K-Nearest Neighbor (K-NN) is a machine learning algorithm that finds and evaluates the closest distance between the data and its neighbors in the classification process [14]. In classification research, In classification research, K-NN is a commonly used method, as demonstrated in the study about determining majors in vocational schools, with 84% accuracy [15]; the stroke disease detection with the highest accuracy value of 95.93% [16]; the flood prediction in Bangladesh city with an accuracy of 94.91% [17]; the non-performing loan prediction with an accuracy of 89.41% [18]; and the rice varieties classification with an accuracy value of 97.15% [19].…”