Muditadevi, Karina and Djunaidi, Karina and Ningrum, Rahma Farah (2024) OPTIMASI K-NEAREST NEIGHBOR (KNN) UNTUK PENCARIAN OBAT BERDASARKAN INDIKASI PENYAKIT. Diploma thesis, ITPLN.
![[thumbnail of 202031030_Karina Muditadevi_Revisi_Skripsi_KARINA Muditadevi.pdf]](https://repository.itpln.ac.id/style/images/fileicons/text.png)
202031030_Karina Muditadevi_Revisi_Skripsi_KARINA Muditadevi.pdf
Restricted to Repository staff only
Download (1MB)
Abstract
Penggunaan obat yang sesuai fungsinya sangat penting untuk memastikan efektivitas pengobatan dan menghindari efek samping yang tidak diinginkan. Penelitian ini bertujuan untuk mengoptimalkan metode K-Nearest Neighbor (KNN) dalam pencarian obat guna meningkatkan akurasi dan efisiensi identifikasi obat yang relevan berdasarkan indikasi tertentu. Metode K-Nearest Neighbor (KNN) dipilih karena kemampuannya mengklasifikasikan nama obat berdasarkan indikasi yang diberikan dan memberikan hasil tetangga terdekat lainnya. Data yang digunakan dalam penelitian ini diperoleh melalui teknik web scraping dari situs Alodokter.com, menghasilkan 1.281 data obat. Pengujian dilakukan menggunakan dua metrik jarak, yaitu Euclidean dan Manhattan, serta teknik Cross-Validation. Hasil menunjukkan bahwa metode K-Nearest Neighbor (KNN) dengan metrik jarak Euclidean mencapai akurasi 82,48%, lebih unggul dibandingkan metrik jarak Manhattan, sehingga memberikan hasil yang lebih relevan dalam pencarian obat berdasarkan indikasi.
The appropriate use of medication is crucial to ensuring the effectiveness of treatment and avoiding undesirable side effects. This study aims to optimize the K Nearest Neighbor (KNN) method in functional drug search to enhance the accuracy and efficiency of identifying relevant drugs based on specific indications. The K-Nearest Neighbor (KNN) method was chosen for its ability to classify drug names based on the given indications and provide results for other nearest neighbors. The data used in this study were obtained through web scraping techniques from the Alodokter.com website, resulting in 1,281 drug entries. The testing was conducted using two distance metrics, Euclidean and Manhattan, as well as the Cross-Validation technique. The results showed that the K-Nearest Neighbor (KNN) method with the Euclidean distance metric achieved an accuracy of 82.48%, outperforming the Manhattan distance metric, thus providing more relevant results in the functional drug search.
Item Type: | Thesis (Diploma) |
---|---|
Uncontrolled Keywords: | K-Nearest Neighbor, Pencarian Fungsi Obat, Euclidean, Manhattan, Cross-Validation K-Nearest Neighbor, Functional Drug Search, Euclidean, Manhattan, Cross-Validation |
Subjects: | Skripsi Bidang Keilmuan > Teknik Informatika |
Divisions: | Fakultas Telematika Energi > S1 Teknik Informatika |
Depositing User: | Sudarman |
Date Deposited: | 15 Sep 2025 02:01 |
Last Modified: | 15 Sep 2025 02:01 |
URI: | https://repository.itpln.ac.id/id/eprint/1043 |