DIAGNOSA PENYAKIT PARKINSON DENGAN ALGORITMA K-NEAREST NEIGHTBOR DAN DECISION TREE C4.5

  • Desiani A
  • Narti N
  • Ramayanti I
  • et al.
N/ACitations
Citations of this article
6Readers
Mendeley users who have this article in their library.

Abstract

Abstrak Parkinson adalah suatu penyakit dimana neurologis mempengaruhi neuron dopaminergik, yang dibuktikan dengan kematian sel-sel otak yang ada, hilangnya pigmentasi substantia nigra, adanya inklusi sitoplasma, dan penurunan kadar dopamin di substantia nigra pars compacta dan corpus striatum. Penyakit parkinson dapat didiagnosa dengan melakukan pengklasifikasian untuk mengukur tingkat akurasi. Tujuan dari penelitian ini adalah untuk melakukan diagnosa penyakit Parkinson dengan dua algoritma yang berbeda, yaitu algoritma K-Nearest Neighbor (KNN) dan algoritma C4.5 dengan metode pelatihan Percentage split dan validasi K-fold cross yang nantinya kan dibandingkan satu sama lain. Dari penelitian ini, nilai presisi yang dimiliki penderita Parkinson's disease algoritma C4.5 split persentasenya adalah 96%. Begitu juga untuk nilai recall yang dimiliki oleh penderita penyakit Parkinson yaitu sebesar 93%. Nilai akurasi algoritma K-Nearest Neighbor (KNN) adalah 82% untuk metode pelatihan pada percentage split dan 76,8% dengan metode validasi K-fold cross dan 89% untuk algoritma C4.5 dengan metode pelatihan pada Percentage split dan 81% dengan metode validasi K-fold cross. Abstract Parkinson's is a disease in which neurological damage affects dopaminergic neurons, as evidenced by the death of existing brain cells, loss of substantia nigra pigmentation, presence of cytoplasmic inclusions, and decreased levels of dopamine in the substantia nigra pars compacta and corpus striatum. Parkinson's disease can be diagnosed by classifying it to measure the level of accuracy. The purpose of this research is to diagnose Parkinson's disease with two different algorithms, namely the K-Nearest Neighbor (KNN) algorithm and the C4.5 algorithm with the split percentage training method and K-fold cross validation which will later be compared with each other. From this study, the percentage of precision that is owned by patients with Parkinson's disease algorithm C4.5 split is 96%. Likewise for the recall value that is owned by people with Parkinson's disease, which is equal to 93%. The accuracy value of the K-Nearest Neighbor (KNN) algorithm is 82% for the training method on the split percentage and 76.8% with the K-fold cross

Cite

CITATION STYLE

APA

Desiani, A., Narti, N., Ramayanti, I., Arhami, M., & Irmeilyana, I. (2023). DIAGNOSA PENYAKIT PARKINSON DENGAN ALGORITMA K-NEAREST NEIGHTBOR DAN DECISION TREE C4.5. Jurnal Simantec, 12(1), 47–58. https://doi.org/10.21107/simantec.v12i1.21167

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free