REKOMENDASI KESEHATAN JANIN DENGAN PENERAPAN ALGORITMA C5.0 MENGGUNAKAN CLASSIFYING CARDIOTOCOGRAPHY DATASET

  • Santoso M
  • Musa P
N/ACitations
Citations of this article
69Readers
Mendeley users who have this article in their library.

Abstract

ABSTRAK Memiliki buah hati yang sehat tidak luput dari faktor kesehatan sang ibu dan kondisi janin di dalam rahim, sehingga butuh analisa terhadap kesehatan janin pada setiap ibu hamil. Penelitian ini mengusulkan Algoritma C5.0 memanfaatkan dataset Cardiotocography terkait kondisi janin. Dataset Cardiotocography terdiri dari 2.126 record dengan, dimana setiap record memiliki 22 kolom atribut dan terdapat 3 kelas klasifikasi, yaitu; normal, suspect, dan pathological. Dengan menghitung entropy, information gain, split information, dan gain ratio, serta menggunakan confusion matrix sebagai perbandingan akurasi dari data yang diteliti. Kelompok record dibagi menjadi data training dan data testing dengan variasi 60%, 70%, 80%, dan 90% untuk data training dan dengan 40%, 30%, 20%, 10% untuk data testing. Hasil dari rekomendari pada variasi pembagian 90% data training dan 10% data testing menghasilkan akurasi sebesar 93,40% dengan jumlah aturan sebanyak 257. Sedangkan variasi 80% data training dan 20% data testing menghasilkan akurasi sebesar 91,29% dengan jumlah aturan sebanyak 239. Pada variasi 70% data training dan 30% data testing menghasilkan akurasi sebesar 88,23% dengan jumlah aturan sebanyak 220. Dan variasi pembagian terkecil 60% data training dan 40% data testing menghasilkan akurasi sebesar 88,12% dengan jumlah aturan sebanyak 204. Berdasarkan variasi tersebut, maka dapat disimpulkan semakin besarnya data training akan menyebabkan akurasi menjadi semakin baik dengan jumlah aturan-aturan yang dapat berguna untuk dijadikan sebagai sistem penunjang keputusan. ABSTRACT Having a healthy baby cannot be separated from the health factors of the mother and the condition of the fetus in the womb, so it is necessary to analyze the fetal health of every pregnant woman. This study proposes the C5.0 Algorithm utilizing the Cardiotocography dataset related to fetal conditions. The Cardiotocography dataset consists of 2,126 records, where each record has 22 attribute columns, and there are three classification classes: normal, suspect, and pathological. By calculating entropy, information gain, split information, and gain ratio, and using confusion matrix to compare the accuracy of the data under study. The record group is divided into training data and testing data with variations of 60%, 70%, 80%, and 90% for training data and with 40%, 30%, 20%, 10% for testing data. The recommendations on the variation 90% of training data and 10% of testing data, which accuracy of 93,40% with 257 rules. While the variation of 80% of training data and 20% of testing data resulted in an accuracy of 91,29% with 239 rules. 70% of training data and 30% of testing data yields an accuracy of 88,23% with several rules of 220. Furthermore, the most negligible distribution variation of 60% training data and 40% testing data results in an accuracy of 88,12% with several rules of 204. Based on these variations, it can be is concluded that the greater the training data, the better the accuracy with the number of rules that can be used as a decision support system.

Cite

CITATION STYLE

APA

Santoso, M. R., & Musa, P. (2021). REKOMENDASI KESEHATAN JANIN DENGAN PENERAPAN ALGORITMA C5.0 MENGGUNAKAN CLASSIFYING CARDIOTOCOGRAPHY DATASET. Jurnal Simantec, 9(2), 65–76. https://doi.org/10.21107/simantec.v9i2.10730

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