Evaluation of Multi-Class Classification Performance Lung Cancer Through K-NN and SVM Approach
Lung cancer is one of the deadliest diseases in the world with a mortality rate of 25% of all cancer-related deaths in 2021. Lung cancer is a lung disease caused by genetic changes in respiratory epithelial cells, resulting in uncontrolled cell proliferation. In an effort to improve diagnosis and tr...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Fakultas Ilmu Komputer UMI
2025-04-01
|
| Series: | Ilkom Jurnal Ilmiah |
| Subjects: | |
| Online Access: | https://jurnal.fikom.umi.ac.id/index.php/ILKOM/article/view/2464 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Lung cancer is one of the deadliest diseases in the world with a mortality rate of 25% of all cancer-related deaths in 2021. Lung cancer is a lung disease caused by genetic changes in respiratory epithelial cells, resulting in uncontrolled cell proliferation. In an effort to improve diagnosis and treatment, this study proposes an approach for multiclass performance evaluation using K-Nearest Neighbors (KNN) and Support Vector Machine (SVM) algorithms based on 2024 data. in this study KNN is implemented conventionally while SVM applies 2 kernel processes, namely Linear and Polynominal. The data used is 1000 rows and uses 24 variables with a ratio of 70% training data and 30% testing data, the data in this study includes important information such as medical history, diagnostic test results, and clinical characteristics of patients. this study aims to determine which algorithm has the best performance by looking at the final results based on accuracy in identifying lung cancer data. Based on the research and discussion of SVM and KNN performance evaluation, the SVM algorithm produces an accuracy of 98.28%, surpassing the accuracy of the KNN algorithm of 97.25%. Therefore, the results show that the SVM algorithm is superior to the KNN algorithm. The KNN and SVM methods were implemented for multi-class classification of lung cancer, allowing identification of various subtypes of lung cancer with optimal accuracy. |
|---|---|
| ISSN: | 2087-1716 2548-7779 |