Skripsi

KOMBINASI METODE MICE DAN ADASYN UNTUK PENANGANAN DATA HILANG DAN KETIDAKSEIMBANGAN DATA PADA KLASIFIKASI PENYAKIT JANTUNG

---

The quality of data is determined by several factors, namely data completeness and data balance. The University of California Irvine (UCI) heart disease dataset has issues with missing data and data imbalance, which, if not addressed, can lead to reduced model prediction accuracy and errors in data interpretation. Missing data can be handled using several methods, one of which is data imputation. For missing data less than or equal to 5%, it can be handled using the mode method for nominal attributes. For missing data greater than 5%, the Multiple Imputation by Chained Equations (MICE) method is used. Data imbalance can be addressed using several methods, one of which is oversampling with Adaptive Synthetic Sampling Approach (ADASYN). To see how handling missing data and data imbalance affects the performance of heart disease classification, performance testing was conducted using the Random Forest, Support Vector Machine (SVM), and Multi-Layer Perceptron (MLP) classification methods. There was an increase in classification performance results after handling missing data and data imbalance. Accuracy increased by 18.48%, precision increased by 18.5%, and recall increased by 18.4%. Based on the obtained results, it can be concluded that the MICE and ADASYN methods can improve classification performance on the UCI heart disease dataset

---

Availability

Detail Information

Series Title

-

Call Number

T1543732024

Publisher

Indralaya : Prodi Ilmu Matematika, Fakultas Matematika Dan Ilmu Pengetahuan Alam Universitas Sriwijaya., 2024

Collation

xiv, 58 hlm.; ilus.; tab, 29 cm.

Language

Indonesia

ISBN/ISSN

-

Classification

519.507

Content Type

Text

Media Type

unmediated

Carrier Type

other (computer)

Edition

-

Subject(s)

Prodi Ilmu Matematika
Matematika Statistikal

Specific Detail Info

-

Statement of Responsibility

SEW

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment