DETEKSI DINI PENYAKIT KANKER PARU DENGAN KOMBINASI METODE KLASIFIKASI ADABOOST DAN RANDOM FOREST

• Main Author: Roy Binsar Sinaga, (Author)
• Format: Book
• Published: 2022-07-12.
• Subjects: Thesis; NonPeerReviewed
• Online Access: Link Metadata
• Notes: http://repository.upnvj.ac.id/19703/1/ABSTRAK.pdf; http://repository.upnvj.ac.id/19703/2/AWAL.pdf; http://repository.upnvj.ac.id/19703/3/BAB%201.pdf; http://repository.upnvj.ac.id/19703/4/BAB%202.pdf; http://repository.upnvj.ac.id/19703/5/BAB%203.pdf; http://repository.upnvj.ac.id/19703/6/BAB%204.pdf; http://repository.upnvj.ac.id/19703/7/BAB%205.pdf; http://repository.upnvj.ac.id/19703/8/DAFTAR%20PUSTAKA.pdf; http://repository.upnvj.ac.id/19703/9/RIWAYAT%20HIDUP.pdf; http://repository.upnvj.ac.id/19703/10/LAMPIRAN.pdf; http://repository.upnvj.ac.id/19703/11/HASIL%20PLAGIARISME.pdf; http://repository.upnvj.ac.id/19703/12/ARTIKEL%20KI.pdf
• Keywords: QA75 Electronic computers. Computer science; T Technology (General)

Lung cancer is the most common type of cancer in men and the fifth most common among other cancers in women in Indonesia. To find out whether a person has lung cancer or not, it is necessary to carry out a series of diagnostic stages by quite a lot of medical personnel and the results cannot be obtained quickly. Therefore, in this study, an experiment will be conducted to use a combination of the Adaboost and Random Forest classification algorithms in early detection of lung cancer cases based on a series of indicators or attributes related to lung cancer. The data used in this study is secondary data from the kaggle.com website. The data consists of 309 data records with 10 features and 1 class which is then preprocessed. Followed by the formation of a model consisting of two models, namely the Random Forest model and the Adaboost model which applies Random Forest as a weak learner. After the model is formed, testing is carried out with test data, where data is divided using hold-out validation twice, namely a scale of 70: 30 and 80: 20. The evaluation results show that the combination of the Adaboost and Random Forest methods produces accuracy, precision, recall, and specificity. the highest were 95.40%, 96%, 96.30%, 96%, respectively. These results are better than the application of the Random Forest method without Adaboost which produces the highest accuracy, precision, recall, and specificity of 93.20%, 92.70%, 96.00%, 95.90%, respectively.