KLASIFIKASI ABNORMALITAS JANTUNG ANAK DENGAN ARSITEKTUR CONVOLUTIONAL NEURAL NETWORKS BINARI DAN MULTI-KELAS
Total View This Week0

Institusion

Universitas Sriwijaya

Author

UNIGHA, SITI LUTHFIA
Nurmaini, Siti

Subject

T57.6.A2-Z General works Simulation Cf. QA76.9.C65 Computer science Cf. TA343 Engineering mathematics 

Datestamp

2023-05-26 02:28:14 

Abstract :
This study aims to develop a classification model using Convolutional Neural Networks (CNN) architecture to identify heart abnormalities in children. In this study, binary and multi-class CNNs are used to process data from children's heart images and produce abnormality class predictions. The data used in this study comes from two categories: normal hearts and hearts with abnormalities. The results of the study show that both CNN models (binary and multi-class) successfully classified children's heart images with a high level of accuracy. The best performance achieved in the case of classifying abnormalities in Infant is by ResNet101 with an accuracy of 94.75% for the abnormality class, while the accuracy for the preview class is 99%. For the unseen data in the view class, the obtained accuracy is 94.2%, and for the unseen data in the abnormality class, the obtained accuracy is 94.75%. In conclusion, the results of this study show that Convolutional Neural Networks architecture can be used to classify heart abnormalities in children with a high level of accuracy. This model can be a useful tool in quickly and accurately diagnosing heart abnormalities in children 

File :
RAMA_56021_09011181924016.pdf
RAMA_56201_09011181924016_TURNITIN.pdf
RAMA_56201_09011281924016_0002085908_01_front_ref.pdf
RAMA_56201_09011281924016_0002085908_02.pdf
RAMA_56201_09011281924016_0002085908_03.pdf
RAMA_56201_09011281924016_0002085908_04.pdf
RAMA_56201_09011281924016_0002085908_05.pdf
RAMA_56201_09011281924016_0002085908_06_ref.pdf
RAMA_56201_09011281924016_0002085908_07_lamp.pdf