Skripsi
PENANGANAN DATA HILANG MENGGUNAKAN SELEKSI FITUR C4.5 PADA KLASIFIKSI PENYAKIT JANTUNG MENGGUNAKAN BACKPROPAGATION
Missing Value can result in weak prediction accuracy, so it needs to be addressed. The way to deal with lost data is to use the deletion method. This method requires a classification algorithm to select the most influential features. In this study, the University of California Irvine (UCI) heart attack dataset was used which has a weakness, namely it has missing data on several features. This study uses the C4.5 Algorithm to overcome missing data by selecting important features and using Backpropagation for testing. The test results obtained an accuracy value of 78.9116 % for testing without the application of the C4.5 algorithm on missing data, while the accuracy value of 80.6112% for testing with the application of the C4.5 algorithm on missing data using 10 features in the test. It can be concluded that the application of the C4.5 algorithm to select influential features and the use of Backpropagation for testing can improve accuracy.
Inventory Code | Barcode | Call Number | Location | Status |
---|---|---|---|---|
2207001020 | T69097 | T690972022 | Central Library (Referens) | Available |
No other version available