Nugraha, Hari (2020) ANALISIS PERBANDINGAN PADA METODE PENGHITUNGAN JARAK ANTAR DATA PADA ALGORITMA K-NN DAN LVQ UNTUK KLASIFIKASI DATA. Tugas Akhir thesis, University of Technology Yogyakarta.
|
Text
Naskah Publikasi-Hari Nugraha-5160411129.pdf Download (530kB) | Preview |
Abstract
Classification is a technique used to build classification models from training data samples. K-NN and LVQ are classification algorithms that have important parameters that affect their performance. The parameter is the value of the distance calculation method. The distance between the two data points is determined by the distance matrix calculation before the classification process is carried out by the two algorithms. The purpose of this study is to analyze and compare the accuracy of K-NN and LVQ performance using the Euclidean Distance, Manhatan Distance, Minkowski Distance, Canberra Distance, Cosine Distance, and Chebishev Distance based on the accuracy point of view. The data used is a dataset sourced from Kaggle.com. The evaluation method used is k-Fold Cross Validation. The analysis shows that Canberra Distance has better performance on both algorithms with an average accuracy of 72.53%. Manhattan Distance has better performance on the k-NN algorithm with an average accuracy of 76.76%. Whereas Cosine Distance has better performance on the LVQ algorithm with an average accuracy of 69.84%. Keywords: Classification, Algorithms, Distance Calculation
| Item Type: | Thesis (Skripsi, Tugas Akhir or Kerja Praktek) (Tugas Akhir) |
|---|---|
| Subjects: | T Technology > T Technology (General) |
| Divisions: | Fakultas Sains Dan Teknologi > S1 Informatika |
| Depositing User: | Kaprodi S1 Informatika UTY |
| Date Deposited: | 27 Mar 2020 05:32 |
| Last Modified: | 27 Mar 2020 05:32 |
| URI: | http://eprints.uty.ac.id/id/eprint/4921 |
Actions (login required)
 |
View Item |
