KOMBINASI ALGORITMA RELIEFF DAN MULTIPLE IMPUTATION BY CHAINED EQUATION (MICE) UNTUK PENANGANAN DATA HILANG PADA DATASET PENYAKIT GINJAL KRONIS
Total View This Week0
Total View This Week0
Institusion
Universitas Sriwijaya
Author
PUTRI, RIZKI EKA
Desiani, Anita
Suprihatin, Bambang
Desiani, Anita
Suprihatin, Bambang
Subject
QA8.9-QA10.3 Computer science. Artificial intelligence. Computational complexity. Data structures (Computer scienc. Mathematical Logic and Formal Languages
Datestamp
2023-05-19 08:12:48
Abstract :
Missing data is a problem where there are empty values or incomplete values in the data. The University of California Irvine (UCI) Machine Learning Repository dataset that records chronic kidney disease patient data is one of the datasets that has a data loss problem. Lost data handling can be done by two methods, namely deletion and imputation. One of the deletion algorithms is ReliefF, while one of the imputation algorithms is Multiple Imputation by Chained Equation (MICE). This study aims to overcome missing data in the UCI chronic kidney disease dataset using a combination of the ReliefF and MICE algorithms. The results of the combination of algorithms were tested using the K-Nearest Neighbor (KNN) method to determine the increase in classification performance. The results of the ReliefF feature selection show that there are 15 features that influence classification with an increase in accuracy of 23,5%. The results shown from the combination of the two methods are an increase in accuracy, precision, and recall of 25,5%, 19,7%, 23,08%. The chronic kidney disease dataset experienced significant increases in accuracy, precision, and recall. It can be concluded that the combination of the ReliefF and MICE algorithms can improve classification performance in the UCI chronic kidney disease dataset.
Missing data is a problem where there are empty values or incomplete values in the data. The University of California Irvine (UCI) Machine Learning Repository dataset that records chronic kidney disease patient data is one of the datasets that has a data loss problem. Lost data handling can be done by two methods, namely deletion and imputation. One of the deletion algorithms is ReliefF, while one of the imputation algorithms is Multiple Imputation by Chained Equation (MICE). This study aims to overcome missing data in the UCI chronic kidney disease dataset using a combination of the ReliefF and MICE algorithms. The results of the combination of algorithms were tested using the K-Nearest Neighbor (KNN) method to determine the increase in classification performance. The results of the ReliefF feature selection show that there are 15 features that influence classification with an increase in accuracy of 23,5%. The results shown from the combination of the two methods are an increase in accuracy, precision, and recall of 25,5%, 19,7%, 23,08%. The chronic kidney disease dataset experienced significant increases in accuracy, precision, and recall. It can be concluded that the combination of the ReliefF and MICE algorithms can improve classification performance in the UCI chronic kidney disease dataset.
File :
RAMA_44201_08011281924051.pdf
RAMA_44201_08011281924051_TURNITIN.pdf
RAMA_44201_08011281924051_0011127702_0026017102_01_front_ref.pdf
RAMA_44201_08011281924051_0011127702_0026017102_02.pdf
RAMA_44201_08011281924051_0011127702_0026017102_03.pdf
RAMA_44201_08011281924051_0011127702_0026017102_04.pdf
RAMA_44201_08011281924051_0011127702_0026017102_05.pdf
RAMA_44201_08011281924051_0011127702_0026017102_06_ref.pdf
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
