Tansig Based MLP Network Cardiac Abnormality
| Autor: | Jailani Abdul Kadir, F.R. Hashim, Syahrull Hi-Fi Syam Ahmad Jamil, Yulni Januar, Nor Sham Hasan, Baharuddin Mustapha |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
0209 industrial biotechnology
Heartbeat Artificial neural network medicine.diagnostic_test Heart malformation business.industry Computer science Activation function Pattern recognition 02 engineering and technology medicine.disease Sudden death 020901 industrial engineering & automation Heart failure Multilayer perceptron 0202 electrical engineering electronic engineering information engineering medicine 020201 artificial intelligence & image processing Artificial intelligence business Electrocardiography |
| Zdroj: | ICCSCE |
| DOI: | 10.1109/iccsce47578.2019.9068588 |
| Popis: | Cardiac disorder can happen to everybody, irrespective of sex, age or race. Family members' history, however, gives a strong indication of the probable risk of middle heart failure. Cardiac anomaly hardly shows early signs, resulting in patient's sudden death. Heartbeat is generally an irregular electrical boost or activity of the heart. In this research, the Multilayer Perceptron (MLP) network is used as an early monitoring system to detect cardiac abnormality. The database of the MIT-BIH is used to extract the cardiac abnormality dataset, which was then used to train the chosen MLP network with multiple training algorithms by using Tansig as the activation function for the MLP network. The research shows the best result given by MLP network using BR training algorithm with 0.0012 on mean square error (MSE) and 0.9955 on regression performance outperform others techniques. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|