Firdaus, Firmantoro (2021) analisis perbandingan antara algoritma k-nearest neighbor dan c4.5 untuk mencari kinerja terbaik dalam penerapan gaya belajar siswa kinestetik & visual (studi kasus smk negeri 1 samarinda). S1 Sistem Informasi thesis, STMIK Widya Cipta Dharma.
![[img]](http://repository.wicida.ac.id/style/images/fileicons/text.png) |
Text
1541101-S1-Jurnal.pdf Restricted to Repository staff only Download (237kB) | Request a copy |
|
Text
1541101-S1-Sistem Informasi.pdf Restricted to Repository staff only Download (1MB) | Request a copy |
Abstract
Tujuan penelitian ini adalah untuk mengetahui hasil akurasi pada algoritma K-Nearest Neighbor dan C4.5 dalam klasifikasi gaya belajar siswa SMK Negeri 1 Samarinda dan mengetahui hasil perbandingan dalam mencari kinerja terbaik dari kedua algoritma tersebut dengan menggunakan uji T-Test. Metode penelitian yang digunakan dalam penelitian ini mengacu pada tahapan proses CRISP-DM (CRoss-Industry Standard Process for Data Mining). Ada 6 tahapan yang harus dilakukan yaitu, Pemahaman Bisnis, Pemahaman Data, Persiapan Data, Pemodelan, Evaluasi, dan Penerapan. Hasil penelitian ini berdasarkan hasil pengujian pada kedua algoritma klasifikasi K-Nearest Neighbor menghasilkan Accuracy 99.49%, Precision 99.50%, dan Recall 98.92%, sedangkan untuk nilai algoritma C4.5 menghasilkan Accuracy 99.66%, Precision 100%, Recall 98.97%. Kemudian untuk nilai AUC pada kedua algoritma memiliki performa yang kurang baik dalam penerapan gaya belajar siswa, untuk algoritma k-NN memiliki nilai AUC 0.500, sedangkan algoritma C4.5 memiliki nilai AUC 0.650. Sedangkan hasil kinerja yang diuji menggunakan T-Test pada algoritma K-Nearest Neighbor dan C4.5 dapat dilihat bahwa algoritma K-Nearest Neighbor memiliki nilai 0.995, sedangkan algoritma C4.5 memiliki nilai tertinggi sebesar 0.997. Dengan demikian algoritma C4.5 dapat memberikan solusi terbaik terhadap penerapan gaya belajar siswa SMK Negeri 1 Samarinda. ========================================================= The purpose of this study is to determine the accuracy of the algorithm K-Nearest Neighbor and C4.5 in the classification of student learning styles of SMK Negeri 1 Samarinda and knowing the results of the comparison in finding the best performance of the two algorithms using the test T-Test. The research method used in this study refers to the stages of the CRISP-DM process ( CRoss-Industry Standard Process for Data Mining). There are 6 stages that must be done, namely, Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation, and Implementation. The results of this study are based on the results of testing on both classification algorithms K-Nearest Neighbor produce Accuracy 99.49%, Precision 99.50%, and Recall 98.92%, while the C4.5 algorithm value yields Accuracy 99.66%, Precision 100%, Recall 98.97%. Then for the AUC value on the two algorithms have a poor performance in the application of student learning styles, for the algorithm k- NN has an AUC value of 0.500, while the C4.5 algorithm has an AUC value of 0.650. While the performance results were tested using T-Test on the algorithm K-Nearest Neighbor and C4.5 it can be seen that the algorithm K-Nearest Neighbor has a value of 0.995, while the C4.5 algorithm has the highest value of 0.997. Thus the C4.5 algorithm can provide the best solution to the application of the learning styles of students of SMK Negeri 1 Samarinda.
| Item Type: | Thesis (S1 Sistem Informasi) |
|---|---|
| Additional Information: | Pembimbing Utama : Salmon, S.Kom., M.Kom Pembimbing Pendamping :H.Pajar Pahrudin, S.Kom., M.H |
| Uncontrolled Keywords: | Data Mining, Classification, K-Nearest Neighbor, C4.5 |
| Subjects: | Q Science > QA Mathematics > QA76 Computer software |
| Divisions: | Sistem Informasi |
| Depositing User: | Mr Firmantoro Firdaus |
| Date Deposited: | 10 Feb 2021 07:17 |
| Last Modified: | 10 Feb 2021 07:17 |
| URI: | http://repository.wicida.ac.id/id/eprint/3563 |
Actions (login required)
 |
View Item |
