Penentuan Pusat Klaster Secara Otomatis Pada Algoritma Density Peaks Clustering Berbasis Metode Inter Quartile Range

Efendi, Ridwan (2024) Penentuan Pusat Klaster Secara Otomatis Pada Algoritma Density Peaks Clustering Berbasis Metode Inter Quartile Range. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 20081010071_Cover.pdf Download (691kB)
	Text (Bab 1) 20081010071_Bab1.pdf Download (210kB)
	Text (Bab 2) 20081010071_Bab2.pdf Restricted to Repository staff only until 30 July 2026. Download (518kB)
	Text (Bab 3) 20081010071_Bab3.pdf Restricted to Repository staff only until 30 July 2026. Download (769kB)
	Text (Bab 4) 20081010071_Bab4.pdf Restricted to Repository staff only until 30 July 2026. Download (2MB)
	Text (Bab 5) 20081010071_Bab5.pdf Download (130kB)
	Text (Daftar pustaka) 20081010071_Daftar_Pustaka.pdf Download (240kB)

Abstract

Clustering is a method for grouping similar data into the same part. This process helps humans to obtain information more quickly. In the context of social media, for example, clustering methods can provide insights into content preferences and dislikes. The Density Peaks Clustering (DPC) algorithm is a popular choice for data clustering. Many studies have utilized this algorithm. However, DPC algorithm suffers from shortcomings in determining cluster centers. Cluster centers in DPC are still manually selected through decision graphs, introducing subjectivity and instability. To address this issue, an algorithm or method called 'Automatic Center Determination' based on the Inter Quartile Range (IQR) method is employed. In this study, the combination of DPC and Automatic Center Determination is referred to as DPC IQRSM algorithm. The research evaluates the effectiveness of the DPC-IQRSM algorithm using low-dimensional datasets such as iris, aggregation, flame, and spiral. Additionally, comparisons are made with two other clustering algorithms, K-means and GB-DPC (Gab Based Density Peak Clustering). The evaluation results indicate that the DPC-IQRSM algorithm outperforms the others.

Item Type:

Thesis (Undergraduate)

Contributors: