Analisis gangguan spektrum autis: penerapan metode supervised learning untuk identifikasi indikator autisme pada individu

Widjaya, Tiffany (2024) Analisis gangguan spektrum autis: penerapan metode supervised learning untuk identifikasi indikator autisme pada individu. Bachelor thesis, Universitas Pelita Harapan.

Preview	Text (Title) Title.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (31kB) | Preview
Preview	Text (Abstract) Abstract.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (254kB) | Preview
Preview	Text (ToC) ToC.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (779kB) | Preview
Preview	Text (Chapter1) Chapter1.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (366kB) | Preview
	Text (Chapter2) Chapter2.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (2MB)
	Text (Chapter3) Chapter3.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (4MB)
	Text (Chapter4) Chapter4.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (2MB)
	Text (Chapter5) Chapter5.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (185kB)
Preview	Text (Bibliography) Bibliography.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (678kB) | Preview
	Text (Appendices) Appendices.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (4MB)

Abstract

Autism Spectrum Disorder (ASD) mendapat perhatian khusus dalam kelainan perkembangan karena pengaruhnya terhadap perilaku neurobehavioral individu. Kesulitan interaksi sosial dan manajemen emosi yang merupakan karakteristik ASD sering membuat pengidap mengalami kesulitan berkomunikasi dan beradaptasi dengan situasi baru. Proses diagnosa autisme yang manual dan subjektif sering kali menyebabkan hasil yang tidak konsisten dan membutuhkan waktu lama, menunjukkan kebutuhan akan solusi alternatif yang lebih efisien dan akurat. Untuk mengatasi masalah ini, telah dikembangkan sebuah model screening autisme menggunakan metode Supervised Learning dengan Support Vector Regression (SVR) dan kernel RBF. Model ini dibangun dari data yang dikumpulkan melalui survei menggunakan metode CARS2-QPC, di mana survei ini disebarkan secara luas untuk mendapatkan data yang representatif dari berbagai demografi. Data yang terkumpul, dengan bantuan dokter spesialis, kemudian dikonversi menjadi format CARS2-ST yang lebih standar. Hasil pelatihan model menunjukkan Mean Squared Error (MSE) sebesar 27.41, Mean Absolute Error (MAE) sebesar 4.54, dan koefisien determinasi (R²) sebesar 0.816, menandakan tingkat akurasi yang tinggi dalam menscreening ASD, sehingga memungkinkan identifikasi lebih dini dan intervensi yang lebih tepat bagi individu dengan ASD. / Autism Spectrum Disorder (ASD) receives special attention within developmental disorders due to its impact on individual neurobehavioral behavior. The characteristics of ASD, including difficulties in social interaction and emotion management, often cause sufferers to struggle with communication and adapting to new situations. The manual and subjective process of autism diagnosis often leads to inconsistent results and is time-consuming, highlighting the need for more efficient and accurate alternative solutions. To address this issue, an autism detection model using Supervised Learning with Support Vector Regression (SVR) and an RBF kernel has been developed. This model is built from data collected through a survey using the CARS2-QPC method, which was widely disseminated to gather representative data from various demographics. With the assistance of a specialist doctor, the collected data was then converted into a more standardized CARS2-ST format. The training results show a Mean Squared Error (MSE) of 27.41, a Mean Absolute Error (MAE) of 4.54, and a coefficient of determination (R²) of 0.816, indicating a high level of accuracy in detecting ASD. This allows for earlier identification and more appropriate interventions for individuals with ASD.

Actions (login required)

View Item