Implementation of ANN optimized by Simulated Annealing in Predicting Side Effect: Case Study Eye Disorders - Dalam bentuk pengganti sidang - Artikel Jurnal
AUZAN FADHLAN

Informasi Umum

Kode

24.04.5684

Klasifikasi

006.312 - Data mining

Jenis

Karya Ilmiah - Skripsi (S1) - Reference

Subjek

Data Mining

Dilihat

28 kali

Informasi Lainnya

Abstraksi

Drug side effects are unwanted responses by the body. The high failure and potential side effects can halt drug production. The absence of tools to transfer information between proteins and organ systems causes to inefficiencies in the drug discovery process. Critical factors in side effects include the drug's chemical substructure and the metabolic pathways of the drug's target. This study utilizes the processed dataset from the SIDER website, dataset contains 1121 records, with 33 duplicates that have been removed. The dataset is split into a 70:30 ratio for training and testing, resulting in 761 training sets and 327 testing sets. The use of machine learning has been proposed as an alternative for predicting drug side effects. While positive results have been achieved, the use of deep learning, particularly artificial neural networks (ANN), has not been extensively explored in predicting drug side effects. The aim of this study is to develop a prediction model for drug side effects in eye disorders using ANN optimized with simulated annealing. The parameter to be optimized is hidden layer, hidden node, activation, and optimizer. Based on the study results, with 1 hidden layer, 149 hidden nodes, activation relu, and optimizer adam, the optimized ANN model developed giving an accuracy value of 0.67 and an F1-score value of 0.67.

Koleksi & Sirkulasi

Tersedia 1 dari total 1 Koleksi

Pengarang

Penerbit

Sirkulasi

Download / Flippingbook