Rohman, Saeful (2026) IMPLEMENTASI SUPERVISED LEARNING MENGGUNAKAN ALGORITMA RANDOM FOREST UNTUK PENYESUAIAN TINGKAT KESULITAN ADAPTIF PADA GAME TERAPI DISLEKSIA BERBASIS ANDROID. S1 / D3 thesis, Universitas Kuningan.

[thumbnail of ABSTRAK] Text (ABSTRAK)
ABSTRAK.pdf

Download (2MB)
[thumbnail of BAB I] Text (BAB I)
BAB I.pdf

Download (277kB)
[thumbnail of BAB II] Text (BAB II)
BAB II.pdf
Restricted to Repository staff only

Download (952kB) | Request a copy
[thumbnail of BAB III] Text (BAB III)
BAB III.pdf
Restricted to Repository staff only

Download (2MB) | Request a copy
[thumbnail of BAB IV] Text (BAB IV)
BAB IV.pdf
Restricted to Repository staff only

Download (1MB) | Request a copy
[thumbnail of BAB V] Text (BAB V)
BAB V.pdf

Download (146kB)
[thumbnail of DAFTAR PUSTAKA] Text (DAFTAR PUSTAKA)
DAFTAR PUSTAKA.pdf

Download (198kB)
[thumbnail of LAMPIRAN] Text (LAMPIRAN)
LAMPIRAN.pdf
Restricted to Repository staff only

Download (6MB) | Request a copy
Official URL: https://rama.uniku.ac.id

Abstract

Terapi konvensional bagi penyandang disleksia umumnya menggunakan media fisik statis yang tidak mampu menyesuaikan tingkat kesulitan secara dinamis, sehingga rentan memicu kebosanan atau frustrasi pada anak. Menindaklanjuti permasalahan tersebut, peneliti mengembangkan aplikasi game terapi disleksia adaptif berbasis Android untuk melatih kemampuan pemrosesan fonologis dan visual secara personal. Peneliti menerapkan pendekatan Supervised Learning menggunakan algoritma Random Forest sebagai mesin pengambil keputusan Dynamic Difficulty Adjustment (DDA). Dalam mengimplementasikan sistem, peneliti mengumpulkan dataset performa dari 20 anak disleksia beserta 3 terapis di Klinik Anak Mitra Cirebon. Peneliti kemudian melatih 100 pohon keputusan untuk memprediksi penyesuaian level secara real-time via ONNX Runtime. Hasil pengujian menunjukkan bahwa model Random Forest mencapai tingkat akurasi sebesar 86,46% dalam mengklasifikasikan performa pemain. Kinerja ini mengindikasikan bahwa model bekerja secara lebih superior dan tangguh saat dikomparasikan dengan pendekatan statis rule-based Fuzzy Sugeno dari sistem sebelumnya. Selain itu, pengujian efektivitas sistem adaptif melalui Uji T-Test mempertegas penurunan tingkat kesalahan (error rate) pemain yang signifikan secara statistik. Meskipun demikian, penelitian ini masih memiliki keterbatasan pada mekanisme antarmuka yang belum mengintegrasikan teknologi pengenalan suara (voice recognition) untuk validasi pelafalan secara mandiri. Sebagai implikasi, penelitian ini memproduksi instrumen intervensi edukasi digital yang memfasilitasi anak disleksia untuk belajar membaca tanpa beban memori berlebih, sekaligus menyajikan laporan pemantauan performa bermain yang objektif bagi para terapis.

Conventional therapy for individuals with dyslexia generally utilizes static physical media unable to dynamically adjust difficulty levels, which often triggers boredom or frustration in children. Addressing this issue, researchers developed an adaptive Android-based dyslexia therapy game to personally train phonological and visual processing skills. Researchers applied a Supervised Learning approach using the Random Forest algorithm as the Dynamic Difficulty Adjustment (DDA) decision engine. In implementing the system, researchers collected a performance dataset from 20 dyslexic children and 3 therapists at Klinik Anak Mitra Cirebon. Researchers then trained 100 decision trees to predict level adjustments in real-time via ONNX Runtime. Testing results show that the Random Forest model achieved an accuracy rate of 86.46% in classifying player performance. This performance indicates that the model operates more superiorly and robustly when compared to the previous static rule-based Fuzzy Sugeno approach. Furthermore, testing the effectiveness of the adaptive system through the T-Test confirms a statistically significant reduction in the players' error rate. Nevertheless, this research still has limitations in the interface mechanism, which has not yet integrated voice recognition technology for independent pronunciation validation. As an implication, this research produced a digital educational intervention instrument that facilitates dyslexic children to learn reading without excessive memory load, while simultaneously providing objective playing performance monitoring reports for therapists.

Item Type: Thesis (S1 / D3)
Uncontrolled Keywords: Game Terapi, Disleksia, Random Forest, Dynamic Difficulty Adjustment, Onnx Runtime, Therapy Game, Dyslexia
Subjects: T Technology > T Technology (General)
Divisions: Fakultas Ilmu Komputer > S1 Teknik Informatika
Depositing User: S.Kom Saeful Rohman
Date Deposited: 21 Jul 2026 07:26
Last Modified: 21 Jul 2026 07:26
URI: https://rama.uniku.ac.id/id/eprint/5797

Actions (login required)

View Item
View Item