Abidin, Muhammad Zaenal (2014) Pelevelan mandiri dengan metode jaringan syaraf tiruan (Learning Vector Quantization) pada game pembelajaran bahasa Arab. Undergraduate thesis, Universitas Islam Negeri Maulana Malik Ibrahim.
|
Text (Fulltext)
10650095.pdf - Accepted Version Available under License Creative Commons Attribution Non-commercial No Derivatives. Download (7MB) | Preview |
Abstract
INDONESIA:
Bahasa arab sangat penting untuk dipelajari. Banyak cara untuk belajar bahasa arab, salah satunya dengan bermain game RPG dengan konten kosa kata bahasa arab. Dalam mebuat game menarik untuk dimainkan game haruslah cerdas, game mampu memberikan tantangan level secara mandiri dan bermanfaat memberikan pendidikan bahasa arab. Untuk mewujudkannya, game perlu artificial intelegence (Kecerdasan buatan). Metode Learning Vector Quantization merupakan metode klasifikasi data yang dapat mengelompokan data game berupa poin kesehatan dan poin skil untuk menentukan level permainan, sehingga game dapat menentukan level secara mandiri. Dari hasil percobaan yang telah dilakukan, Algoritma Learning Vector Quantization telah dapat diterapkan. Nilai bobot awal dan learning rate dapat mempengaruhi margin error yang diperoleh. Sedangkan nilai epoch juga dapat mempengaruhi hasil akhir sehingga dapat disimpulkan bahwa untuk mendapatkan hasil training data yang akurat harus dilakukan trial dan error sampai di dapatkan hasil yang terbaik. Hasil uji coba digunkan learning rate = 0,1 dengan epoch = 10 untuk memperoleh hasil maksimal, pada proses training dapat diketahui 80% yang berhasil dan 10% yang gagal dalam melakukan proses. Sehingga mampu mengklasifikasi level permainan secara otomatis (automatic leveling) untuk batas–batas perpindahan level pada game pembelajaran bahas arab.
ENGLISH:
The Arabic language is very important to learn. Many ways to learn Arabic, one of them by playing games RPG with Arabic vocabulary content. In making the game interesting to play, the game must be intelligent, able to provide a challenge level of the game independently and helpful providing Arabic language education. To realize, the game needs to artificial Intelligence (AI). Learning Vector Quantization is a data classification method that can classify the game data in the form of health points and skill points to determine the level of play, so the game can determine independently level. From the results of experiments that have been carried, Learning Vector Quantization algorithm has been applied. Initial weight values and the learning rate can affect the margin of error is obtained. While the value of the epoch also can affect the final result, so it can be concluded that the results of training to get accurate data to do trial and error to get the best results on. The trial results digunkan learning rate = 0.1 with epoch = 10 to obtain maximum results, the training process can be seen 80% successful and 10% failed in the process. So as to classify the level of the game automatically (automatic leveling) to limit-the limit on the level of displacement discussed arabic learning games.
Item Type: | Thesis (Undergraduate) | |||||||||
---|---|---|---|---|---|---|---|---|---|---|
Supervisor: | Nurhayati, Hani and Kurniawan, Fachrul | |||||||||
Contributors: |
|
|||||||||
Keywords: | Pelevelan Mandiri; Jaringan Syaraf Tiruan; Bahasa Arab; Autonomic Leveling; Arabic; Neural Network; Game RPG; Learning Vector Quantization | |||||||||
Departement: | Fakultas Sains dan Teknologi > Jurusan Teknik Informatika | |||||||||
Depositing User: | Dian Anesti | |||||||||
Date Deposited: | 13 Oct 2017 14:07 | |||||||||
Last Modified: | 13 Oct 2017 14:07 | |||||||||
URI: | http://etheses.uin-malang.ac.id/id/eprint/8062 |
Downloads
Downloads per month over past year
Actions (login required)
![]() |
View Item |