Skripsi
KLASIFIKASI KELAINAN IRAMA JANTUNG ARITMIA DENGAN KOMBINASI METODE DENOISING AUTOENCODER-AUTOENCODER DAN DEEP NEURAL NETWORK
Arrhythmia is a heart disease caused by abnormal electrical impulses in the heart or abnormalities of the electrocardiogram wave (ECG). In diagnosing it, using a wave from ECG is used in terms of analysis and diagnosis of a heart disease. However, based on the challenges in research and classification of the ECG signal itself, namely the cleanliness of the ECG data from noise. The stages of denoising to classification using deep learning in ECG signals have been used in the last decades and developed using various architectures. By using a combination of Denoising Autoencoder (DAE) as signal denoising, Autoencoder (AE) as feature reduction, and Deep Neural Network (DNN) as an ECG beat classifier. In the first process, the data is cleaned using DAE by previously using Discrete Wavelet Transform (DWT) to create the target signal as a reference during model development. Furthermore, feature reduction is carried out using AE by reducing data that has a feature length of 181 to 91. After the data is reduced, a classification test is carried out using DNN with the 10 test models using various data combination. Based on the test, the best model is model-10 with the results of accuracy, sensitivity, specificity, precision, and F1-Score of 99.77%, 98.94%, 99.71%, 95.76%, 97.25, respectively. Then the unseen test was carried out with different records from the modeling data with the results of accuracy, sensitivity, specificity, precision, and F1-Score of 91.50%, 94.63%, 94.63%, 85.48%, 88.67%, respectively.
Inventory Code | Barcode | Call Number | Location | Status |
---|---|---|---|---|
2107002421 | T49526 | T495262021 | Central Library (Referens) | Available but not for loan - Not for Loan |
No other version available