Search for collections on Undip Repository

Sistem Pendukung Keputusan Penjurusan SMK Berdasarkan Psikotes Bakat Minat Menggunakan Algoritma Backpropagation For Multi Label Learning

ADIKHRESNA, Oxapisi Vidyandika and Kusumaningrum, Retno and Warsito, Budi (2019) Sistem Pendukung Keputusan Penjurusan SMK Berdasarkan Psikotes Bakat Minat Menggunakan Algoritma Backpropagation For Multi Label Learning. Masters thesis, School of Postgraduate Studies.

[img] Text
COVER.pdf

Download (927kB)
[img] Text
BAB I.pdf

Download (214kB)
[img] Text
BAB II.pdf

Download (464kB)
[img] Text
BAB III.pdf
Restricted to Repository staff only

Download (518kB)
[img] Text
BAB IV.pdf
Restricted to Repository staff only

Download (2MB)
[img] Text
BAB V.pdf

Download (227kB)
[img] Text
DAFTAR PUSTAKA.pdf

Download (194kB)

Abstract

Banyak lulusan SMP yang memilih untuk melanjutkan jenjang pendidikannya di Sekolah Menengah Kejuruan (SMK). Menentukan jurusan SMK berarti menentukan bidang karir pula, sehingga pemilihan jurusan adalah pengambilan keputusan yang krusial. Dalam bidang psikologi dikenal Psikotes Bakat Minat, yaitu serangkaian tes psikologi yang mampu menilai aspek-aspek utama kecerdasan, tipe kepribadian, dan bidang minat sesorang. Psikotes ini cocok untuk digunakan sebagai tolak ukur dalam memprediksi jurusan program keahlian SMK yang tepat untuk para calon siswa. Algoritma Backpropagation for Multi Label Learning (BP-MLL) digunakan untuk membuat model untuk menentukan rekomendasi jurusan-jurusan SMK yang cocok berdasarkan hasil psikotes bakat minat calon siswa, kemudian model tersebut digunakan untuk membangun sebuah sistem pendukung keputusan. Hasil penelitian menggunakan 2387 data latih menunjukkan bahwa model jaringan BP-MLL menghasilkan performa nilai hamming loss sebesar 0,30. Berbeda dengan penelitian multi label pada umumnya, pada penelitian ini data latih yang digunakan hanya memiliki 1 label namun tetap mampu memberikan keluaran multi label pada proses prediksi. Model yang sama digunakan untuk melatih data latih dengan multi label dan menghasilkan performa nilai hamming loss sebesar 0,16, walaupun lebih baik namun tidak terpaut jauh. Hal ini menunjukkan bahwa algoritma BP-MLL yang dilatih dengan 1 label mampu menghasilkan performa yang hampir sama dengan yang dilatih dengan multi label.
Kata Kunci : Backpropagation for Multi Label Learning, Penjurusan SMK, Psikotes Bakat Minat

Many junior high school graduates choose to continue their education at Vocational High Schools (SMK). Determining vocational majors means determining career field as well, so the selection of majors is crucial decision making. In the field of psychology, Personality-Aptitude Test is a series of psychological tests to assess the main aspects of intelligence, personality types, and areas of interest. This psychological test is suitable to be used as a benchmark in predicting vocational majors that are appropriate for the prospective students.The Backpropagation for Multi Label Learning (BP-MLL) algorithm is used to create a model for determining suitable vocational majors recommendations based on the results of prospective student's personality-aptitude test, then the model is used to build a decision support system. The results of this study using 2387 training data showed that the BP-MLL network model produced hamming loss value of 0.30. In contrast to multi-label research in general, in this study the training data used only had 1 label but was still able to provide multi-label output in the prediction process. The same model is used to train the training data with multi label and produced hamming loss value of 0.16, although it is better but not far adrift. This shows that the BP-MLL algorithm that was trained with 1 label was able to produce almost the same performance as those trained with multi label.
Keywords : Backpropagation for Multi Label Learning, Vocational High School Majoring, Personality-Aptitude Test

Item Type: Thesis (Masters)
Uncontrolled Keywords: Backpropagation for Multi Label Learning, Penjurusan SMK, Psikotes Bakat Minat
Subjects: Sciences and Mathemathic
Divisions: Postgraduate Program > Master Program in Information System
Depositing User: ekana listianawati
Date Deposited: 18 Apr 2022 04:23
Last Modified: 18 Apr 2022 04:23
URI: https://eprints2.undip.ac.id/id/eprint/5799

Actions (login required)

View Item View Item