EFFECTIVE DIAGNOSIS OF HEART DISEASE PRESENCE USING ARTIFICIAL NEURAL NETWORKS




Abstract:
Due to high complexity of decision making in medicine, it has been proven that usage of Neural Networks is in the cope with the aforementioned problem. Regarding the variety of the symptoms, one of the biggest challenges is heart disease. This research has shown that, depending on the symptoms, Multilayer Perceptron Classifier can effectively decide whether the patient is suffering from heart disease or not. Main goal of this paper is to determine the proper parameters setting for the Multilayer Perceptron algorithm in order to predict heart disease with higher accuracy. However, in order to compare the obtained results using MLP, the experiment is also done using kNN, and LDA algorithms. The results confirm that recognition rate of 96.67%, when using MLP, outperforms other methods when processing heart disease data.

CITATION:

IEEE format

S. Cako, A. Njeguš, V. Matić, “Effective Diagnosis of Heart Disease Presence Using Artificial Neural Networks,” in Sinteza 2017 - International Scientific Conference on Information Technology and Data Related Research, Belgrade, Singidunum University, Serbia, 2017, pp. 3-8. doi:10.15308/Sinteza-2017-3-8

APA format

Cako, S., Njeguš, A., Matić, V. (2017). Effective Diagnosis of Heart Disease Presence Using Artificial Neural Networks. Paper presented at Sinteza 2017 - International Scientific Conference on Information Technology and Data Related Research. doi:10.15308/Sinteza-2017-3-8

BibTeX format
Download

RefWorks Tagged format
Download