Trianto, . (2022) PENERAPAN BORDERLINE-SMOTE DAN GRID SEARCH PADA BAGGING-SVM UNTUK KLASIFIKASI PENYAKIT DIABETES. Skripsi thesis, Universitas Pembangunan Nasional Veteran Jakarta.
![[img]](http://repository.upnvj.ac.id/style/images/fileicons/text.png) |
Text
ABSTRAK.pdf Download (11kB) |
|
Text
AWAL.pdf Download (610kB) |
|
Text
BAB 1.pdf Download (89kB) |
|
Text
BAB 2.pdf Restricted to Repository UPNVJ Only Download (223kB) |
|
Text
BAB 3.pdf Restricted to Repository UPNVJ Only Download (165kB) |
|
Text
BAB 4.pdf Restricted to Repository UPNVJ Only Download (565kB) |
|
Text
BAB 5.pdf Download (77kB) |
|
Text
DAFTAR PUSTAKA.pdf Download (145kB) |
|
Text
RIWAYAT HIDUP.pdf Restricted to Repository UPNVJ Only Download (17kB) |
|
Text
LAMPIRAN.pdf Restricted to Repository UPNVJ Only Download (1MB) |
|
Text
HASIL PLAGIARISME.pdf Restricted to Repository staff only Download (1MB) |
|
Text
ARTIKEL KI.pdf Restricted to Repository staff only Download (522kB) |
Abstract
Diabetes has become a fairly dangerous disease that can cause paralysis to threaten the life of the sufferer. Indonesia is ranked 7th among the 10 countries with the highest number of sufferers with 10.7 million sufferers out of 172.2 million total adult population in 2019. One of the preventive steps that can be taken to avoid the dangers of diabetes is to predict which can predict diabetes by utilizing data mining to build a classification model. One of the classification algorithms that has been widely used in research to detect diabetes is a support vector machine. Although it gives good generalization results, the SVM algorithm has weaknesses when given data with unbalanced classes and it is difficult to determine optimal parameters. To overcome the lack of classification on unbalanced data, we can use the borderline-SMOTE oversampling method which will increase the amount of data in the minor class so that the class distribution becomes balanced. In parameter optimization problems, you can use the grid search method. Then also apply the bagging algorithm to get better classification results by avoiding overfitting and reducing variance. The results showed that the model formed with the SVM, bagging, borderline-SMOTE and grid search algorithms got an accuracy of 92,1%, a precision value of 95,51% for a healthy class and 86,12% for a diabetes class, a recall value of 92,32% for a healthy class and 91,66 % for diabetes class, and f1-score value of 93,39% for healthy class and 88,81% for diabetes class.
| Item Type: | Thesis (Skripsi) |
|---|---|
| Additional Information: | [No.Panggil]: 1810511074 [Pembimbing]: Anita Muliawati [Pembimbing]: Helena Nurramdhani Irmanda [Penguji 1]: Yuni Widiastiwi [Penguji 2]: Mayanda Mega Santoni |
| Uncontrolled Keywords: | classification, diabetes, svm, bagging, borderline-smote, grid search |
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
| Divisions: | Fakultas Ilmu Komputer > Program Studi Informatika (S1) |
| Depositing User: | Trianto - |
| Date Deposited: | 01 Aug 2022 04:46 |
| Last Modified: | 01 Aug 2022 04:46 |
| URI: | http://repository.upnvj.ac.id/id/eprint/19576 |
Actions (login required)
 |
View Item |