Analisis Performansi Metode Jaringan Syaraf Tiruan Learning Vector Quantization Untuk Pendetekisan Anak Berbakat (Gifted Child) Pada Masa Perkembangan

Andyani, Nuri Insania (2014) Analisis Performansi Metode Jaringan Syaraf Tiruan Learning Vector Quantization Untuk Pendetekisan Anak Berbakat (Gifted Child) Pada Masa Perkembangan. Other thesis, Universitas Komputer Indonesia.

[thumbnail of 1.UNIKOM_NURI INSANIA ANDYANI_cover.pdf] Text
1.UNIKOM_NURI INSANIA ANDYANI_cover.pdf

Download (50kB)
[thumbnail of 5.UNIKOM_NURI INSANIA ANDYANI_daftar isi.pdf] Text
5.UNIKOM_NURI INSANIA ANDYANI_daftar isi.pdf

Download (16kB)
[thumbnail of 10.UNIKOM_NURI INSANIA ANDYANI_daftar pustaka.pdf] Text
10.UNIKOM_NURI INSANIA ANDYANI_daftar pustaka.pdf

Download (91kB)
[thumbnail of 4.UNIKOM_NURI INSANIA ANDYANI_kata pengantar.pdf] Text
4.UNIKOM_NURI INSANIA ANDYANI_kata pengantar.pdf

Download (82kB)
[thumbnail of 11.UNIKOM_NURI INSANIA ANDYANI_Bab I.pdf] Text
11.UNIKOM_NURI INSANIA ANDYANI_Bab I.pdf

Download (139kB)
[thumbnail of 12.UNIKOM_NURI INSANIA ANDYANI_Bab II.pdf] Text
12.UNIKOM_NURI INSANIA ANDYANI_Bab II.pdf

Download (875kB)
[thumbnail of 13.UNIKOM_NURI INSANIA ANDYANI_Bab III.pdf] Text
13.UNIKOM_NURI INSANIA ANDYANI_Bab III.pdf

Download (3MB)
[thumbnail of 14.UNIKOM_NURI INSANIA ANDYANI_Bab IV.pdf] Text
14.UNIKOM_NURI INSANIA ANDYANI_Bab IV.pdf

Download (486kB)
[thumbnail of 15.UNIKOM_NURI INSANIA ANDYANI_Bab V.pdf] Text
15.UNIKOM_NURI INSANIA ANDYANI_Bab V.pdf

Download (68kB)
[thumbnail of 18.UNIKOM_NURI INSANIA ANDYANI_lembar pengesahan.pdf] Text
18.UNIKOM_NURI INSANIA ANDYANI_lembar pengesahan.pdf

Download (219kB)
[thumbnail of 23.UNIKOM_NURI INSANIA ANDYANI_surat hak eksklusif.pdf] Text
23.UNIKOM_NURI INSANIA ANDYANI_surat hak eksklusif.pdf

Download (114kB)

Abstract

Anak berbakat(gifted child) adalah anak yang memiliki potensi kecerdasan, kreatifitas dan tanggung jawab terhadap tugas diatas anak�anak seusianya. Setiap anak memiliki gejala-gejala berdasarkan perilaku dan sosial emosinya. Gejala antara satu tipe anak berbakat(gifted child) beririsan dengan tipe anak berbakat(gifted child) yang lainnya. Hal ini dapat menyebabkan proses deteksi menjadi sulit karena ada beberapa gejala yang mirip antara satu tipe dengan tipe yang lainnya. Untuk mengatasi hal tersebut, maka digunakan sebuah metode yaitu metode learning vector quantization yang memiliki kemampuan untuk belajar mengelompokkan pola-pola ke dalam kelas�kelas pola. Data masukan untuk pendeteksian anak berbakat(gifted child) adalah data gejala�gejala berdasarkan perilaku serta sosial emosi anak yang kemudian disusun menjadi 75 buah variabel masukan. Dari gejala�gejala tersebut maka hasil yang diperoleh yaitu tipe anak berbakat(gifted child) yang terdeteksi, persentasi keakuratan terhadap kelas asli dengan kelas hasil deteksi serta optimasi nilai parameter yang bervariasi dan waktu pemrosesan. Adapun tipe anak berbakat(gifted child) yang terdeteksi yaitu tipe the successfuls, the challanging, the underground, the dropouts, the double labeled dan the autonomous learner. Berdasarkan hasil proses pengujian yang menggunakan cross validation dan confusion matrix dengan 5 fold dari jumlah data sebanyak 50 data, maka performansi metode learning vector quantization untuk pendeteksian anak berbakat(gifted child) pada masa perkembangan termasuk performansi yang baik dengan nilai parameter yang optimal pada maksimal epoh = 100, learning rate = 0,02 dan error minimum = 0,0001 dan waktu pemrosesan selama 15 detik memperoleh persentasi keakuratan yang mencapai 100%.

Gifted child is a child who have a potential of intelligence, creativity and responsibility over his/her age. Every child has the symptoms based on their behaviour and social emotion. The symptoms between one type of gifted child have the symptoms that intersect with the other type of gifted child. It results detection process to be difficult, because there is some similar symptoms between one type of gifted child and the other ones. That problem is solved by learning vector quantization method, which has the ability to classify the patterns. Input data for detecting the gifted children are the data of symptoms based on their behavior and social emotion, which compiled to be 75 variables. It produces a type of gifted child who detected, the percentage accuracy between the original class and the detection class, the variation of parameter value and the processing time optimization. As for the type of gifted children who detected are the successfuls, the challanging, the underground, the dropouts, the double labeled and the autonomous learner. According to the results of testing process that uses cross validation and confusion matrix with 5 folds from the amount of data as much as 50 data, so learning vector quantization performance for detection of gifted child in growth fase including the good performance with parameter value optimization on maximum epoh = 100, learning rate = 0,02, error minimum = 0,0001 and the processing time during 15 seconds obtain the percentage accuracy that achieves 100%.

Item Type: Thesis (Other)
Additional Information: Migrasi dari GDL. Identifier: jbptunikompp-gdl-nuriinsani-35459. Folder: S1-Final_Project / Fakultas_Teknik_dan_Ilmu_Komputer / Teknik_Informatika / 2014
Uncontrolled Keywords: Analisis Performansi, Jaringan Syaraf Tiruan, Learning Vector Quantization, Anak Berbakat
Divisions: S1 Skripsi > Teknik Informatika
Depositing User: Robot Migration
Date Deposited: 05 Nov 2026 11:47
Last Modified: 05 Nov 2026 11:47
URI: https://elibrary.unikom.ac.id/id/eprint/45703

Actions (login required)

View Item
View Item