Analisis Komparatif Evaluasi Performa Algoritma Klasifikasi pada Readmisi Pasien Diabetes

Abstract

Abstract. Readmission is associated with quality measures on patients in PendahuluanReadmisi Rumah Sakit (Hospital Readmission) adalah suatu tindakan atau kejadian seorang pasien dirawat kembali yang sebelumnya telah mendapatkan layanan rawat inap dirumah sakit (Iskandar, 2014). Proses readmisi dikaitkan dengan perhitungan kualitas penanganan pasien di rumah sakit (Rubin, 2015;Dungan, 2012). Penyakit diabetes merupakan penyakit metabolisme tubuh yang saat ini belum ada obat yang dapat menyembuhkannya dan dapat menimbulkan penyakit komplikasi-komplikasi lainnya. Pasien penderita diabetes dituntut untuk menjaga tingkat glukosa dalam tubuh agar tetap berada di level yang sehat. Karena tingkat perawatan inap pasien diabetes berada di angka yang tinggi (Ramirez, dkk., 2012), sehingga biaya perawatan yang dibayarkan cenderung mahal (American Diabetes Association, 2013).Perhitungan kualitas terhadap pelayanan penyakit diabetes dapat dilihat dari proses medikasi, etnis, ras, gaya hidup, umur, dll. Di dalam penyakit diabetes, perbedaan atribut terkait pasiennya sangat berpengaruh terhadap kualitas pengobatan yang merujuk pada ketahanan

Comparison of Naïve Bayes Algorithm and Decision Tree C4.5 for Hospital Readmission Diabetes Patients using HbA1c Measurement

“…The confusion matrix of the best results is in Scenario 8 C4.5 on Table 9, it shows the detail data that is successfully classified correctly and the amount of data that is incorrectly classified. From the results of the confusion, the matrix can also be calculated the values of evaluation metrics using (6) to (8) for binary class classification (9) to (11) for multiclass classification. The results of the evaluation metrics of Scenario 8 C4.5 on Table 9 as the best results show an accuracy of 82.74 %, a precision of 87.1 % and a recall of 82.7 %.…”

“…Consequently, the better quality of treatment for the hospital identified by the longer the glycemic serum is at a healthy level. But the differences in attributes associated with diabetic patients result in the calculation of quality, tend to be complicated [8]. The readmission process is very important to anticipate diabetic patients who are late in re-treating their disease.…”

“…The disadvantages of Logistic Regression include vulnerability to underfitting in the imbalance data set and, consequently the value of accuracy is uncertain [12]. Other studies of the classification of Hospital Readmission Diabetes Patients, are compared to Decision Tree algorithms, K-Nearest Neighbor (k-NN), and Naive Bayes with various parameters [8], resulted in the Naïve Bayes classification model having better statistics than other algorithm models such as Decision Tree and k-NN with an accuracy value of 57.52 %, MAE of 0.512, and the kappa statistic of 0.182. There is another study by implementing the C4.5 algorithm to classify the readmissions of diabetic patients, tested the C4.5 algorithm with several different experiments.…”

Penerapan Metode Klasifikasi K-Nearest Neigbor pada Dataset Penderita Penyakit Diabetes

“…Akurasi didefinisikan sebagai tingkat kedekatan antara nilai prediksi dengan nilai actual [5]. rumus akurasi dipaparkan pada persamaan 2.…”

Discrete Wavelet Transform (DWT) and Random Forest for Cancer Detection Based on Microarray Data Classification

“…is the node, and is the probability is in class , and m is the class. If the data is broken down to attribute A into two subsets 1 and 2, then the Gini index equation is as follows ( 5) [19].…”