Klasifikasi Penyakit Malaria Menggunakan Metode Gabungan Vector Quantized Variational Autoencoder Dan Convolutional Neural Networks.

Hidayatullah, Muh Syafiq (2022) Klasifikasi Penyakit Malaria Menggunakan Metode Gabungan Vector Quantized Variational Autoencoder Dan Convolutional Neural Networks. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06111840000096-UNDERGRADUATE_THESIS.pdf Restricted to Repository staff only Download (8MB) | Request a copy

Abstract

Perkembangan penelitian terkait Symbolic AI yang berfokus pada topik Explainable AI dapat memberikan representasi vektor laten yang baik pada metode deep learning. Disisi lain, kasus penyakit malaria masih melebihi angka puluhan ribu di Indonesia. Metode deep learning dapat digunakan untuk mengenali seseorang penderita penyakit malaria atau tidak dengan akurat melalui citra sel darahnya. Akan tetapi, data citra sel darah penderita malaria sejauh ini masih sedikit jika diimplementasikan pada metode deep learning. Oleh karena itu, penelitian ini bertujuan untuk mengkonstruksi citra sel darah penderita malaria berdasarkan vektor laten diskrit untuk proses klasifikasi penderita penyakit malaria. Penelitian ini mengimplementasikan metode gabungan Vector Quantization - Variational Autoencoder (VQ-VAE) sebagai kontruksi citra sel darah dan Convolutional Neural Network (CNN) sebagai proses klasifikasi. Metode gabungan VQ-VAE dan CNN pada penelitian ini dapat memberikan hasil akurasi klasifikasi yang lebih baik sebesar 95.68% dibandingkan metode klasifikasi tanpa konstruksi data dengan akurasi sebesar 94.07%.
==============================================================================================================================
The development of research related to Symbolic AI that focuses on the topic of Explainable AI can provide a good latent vector representation in deep learning methods. On the other hand, malaria cases still exceed tens of thousands in Indonesia. The deep learning method can be used to identify someone with malaria or inaccurately through the image of their blood cells. However, so far, there is very little data on blood cell image data for malaria sufferers if it is implemented in deep learning methods. Therefore, this study aims to construct a discrete latent vector malaria patient blood cell image for the classification process of malaria sufferers. This research implements the combined method of Vector Quantization - Variational Autoencoder (VQ-VAE) as a blood cell image construction and Convolutional Neural Network (CNN) as a classification process. The combined VQ-VAE and CNN methods in this study can provide a better classification accuracy result of 95.68% than the classification method without data construction with an accuracy of 94.07%.

Item Type:	Thesis (Other)
Additional Information:	RSMa 519.53 Hid k-1 2022
Uncontrolled Keywords:	Vektor laten diskrit, Penyakit Malaria, VQ-VAE, CNN. Discrete latent vector, Malaria Disease, VQ-VAE, CNN.
Subjects:	Q Science > QA Mathematics
Divisions:	Faculty of Mathematics and Science > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Mr. Marsudiyana -
Date Deposited:	09 Jun 2026 01:33
Last Modified:	09 Jun 2026 01:33
URI:	http://repository.its.ac.id/id/eprint/133646

Actions (login required)

View Item