Rahmawati, Amalia (2019) Perbandingan Klasifikasi Menggunakan Metode Backpropagation Dan Metode Learning Vector Quantization (Studi Kasus Status Gizi Balita Kabupaten/Kota di Indonesia 2016). Sarjana thesis, Universitas Brawijaya.
Abstract
Klasifikasi merupakan proses menemukan sekumpulan model yang dapat digunakan untuk menggambarkan dan membedakan konsep atau kelas-kelas data yang kelasnya tidak diketahui sebelumnya. Salah satu metode klasifikasi yang berkembang dari kelompok machine learning di bidang artificial intelligence adalah jaringan syaraf tiruan. Penelitian ini bertujuan untuk melakukan perbandingan ketepatan hasil klasifikasi metode backpropagation dan metode learning vector quantization terhadap data status gizi balita kabupaten/kota di Indonesia pada tahun 2016. Ketepatan klasifikasi dapat ditentukan melalui hit ratio, semakin besar hit ratio menunjukkan bahwa hasil klasifikasi semakin baik. Dengan menggunakan 80% data training dan 20% data testing maka diperoleh hit ratio metode backpropagation dengan enam neuron hidden layer sebesar 77,88% dan metode learning vector quantization dengan size codebook 9 sebesar 80,77%. Secara keseluruhan dapat disimpulkan bahwa ketepatan klasifikasi terbaik pada data status gizi balita kabupaten/kota di Indonesia tahun 2016 diperoleh menggunakan metode learning vector quantization dengan size codebook 9.
English Abstract
Classification is the process of finding a set of models that can be used to describe and define concepts or classes of data whose classes are not understood beforehand. One classification method that developed from the machine learning group in the field of artificial intelligence is artificial neural networks. This study aims to compare the accuracy of the results of the classification of the backpropagation method and the learning vector quantization method to the 2016 nutritional data of district / city toddlers in Indonesia. The classification accuracy can be determined through the hit ratio, the greater the hit ratio shows that the classification results are better. By using 80% training data and 20% testing data the hit ratio backpropagation method with siz neuron hidden layers is 77.88% and the learning vector quantization method with size codebook 9 is 80.77%. Overall it can be concluded that the accuracy of the best classification on the nutritional status data of district/city toddlers in Indonesia in 2016 was obtained using the learning vector quantization method with size codebook 9.
Other obstract
-
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/MIPA/2019/63/051910765 |
| Uncontrolled Keywords: | Klasifikasi, Status gizi balita, Backpropagation, Learning vector quantization, Hit ratio. Classification, Nutritional status of children under five years, Backpropagation, Learning vector quantization, Hit ratio. |
| Subjects: | 500 Natural sciences and mathematics > 519 Probabilities and applied mathematics > 519.7 Programming |
| Divisions: | Fakultas Matematika dan Ilmu Pengetahuan Alam > Statistika |
| Depositing User: | Budi Wahyono Wahyono |
| Date Deposited: | 25 Aug 2020 02:40 |
| Last Modified: | 27 Oct 2021 07:35 |
| URI: | http://repository.ub.ac.id/id/eprint/176820 |
Preview |
Text
Amalia Rahmawati (3).pdf Download (2MB) | Preview |
Actions (login required)
 |
View Item |