Optimasi Pusat Klaster K-Prototypes Pada Pengelompokan Penerimaan Bantuan Rehabilitasi Rutilahu di Kota Surabaya Tahun 2024

Ningrum, Lisya Septyo (2025) Optimasi Pusat Klaster K-Prototypes Pada Pengelompokan Penerimaan Bantuan Rehabilitasi Rutilahu di Kota Surabaya Tahun 2024. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 21083010003 Cover.pdf Download (1MB)
	Text (Bab I Pendahuluan) 21083010003 Bab I Pendahuluan.pdf Download (347kB)
	Text (Bab II Tinjauan Pustaka) 21083010003 Bab II Tinjauan Pustaka.pdf Restricted to Registered users only until 2028. Download (591kB)
	Text (Bab III Desain dan Implementasi Sistem) 21083010003 Bab III Desain dan Implementasi Sistem.pdf Restricted to Registered users only until 2028. Download (788kB)
	Text (Bab IV Hasil dan Pembahasan) Bab IV Hasil dan Pembahasan.pdf Restricted to Registered users only until 2028. Download (4MB)
	Text (Bab V Penutup) 21083010003 Bab V Penutup.pdf Download (213kB)
	Text (Daftar Pustaka) 21083010003 Daftar Pustaka.pdf Download (197kB)
	Text (Lampiran) 21083010003 Lampiran.pdf Restricted to Registered users only until 2028. Download (1MB)

Abstract

The issue of substandard housing, known as Rumah Tidak Layak Huni (Rutilahu), represents a significant social challenge that directly impacts the quality of life for low-income communities. In Surabaya, the number of Rutilahu rehabilitation aid applications has increased annually, reflecting the urgent public demand for adequate housing. However, the current selection process for aid recipients still faces major challenges, primarily due to subjectivity in assessment. This subjectivity arises from the absence of an objectively standardized system, often leading to misallocation of aid and ineffective targeting. Therefore, a data-driven, objective, and efficient approach is required to cluster potential beneficiaries in a fair and transparent manner. This study proposes a clustering approach using the K-Prototypes algorithm, which is suitable for handling mixed-type data, to group applicants for the Rutilahu rehabilitation aid program. To address the sensitivity of K-Prototypes to initial cluster centroid selection, three metaheuristic optimization algorithms are Particle Swarm Optimization (PSO), Genetic Algorithm (GA), and Flower Pollination Algorithm (FPA) were employed to enhance clustering performance. The models were evaluated using the Davies-Bouldin Index (DBI), Silhouette Score, and computational time. The results indicate that under the non-weighted clustering scenario, the K-Prototypes model optimized with PSO achieved the best performance, with a DBI of 0,6467, Silhouette Score of 0,5498, and an average computational time of 18,0264 seconds. In the weighted clustering scenario, the conventional K-Prototypes model yielded a slightly better cluster quality with a DBI of 0,5828 and Silhouette Score of 0,5624, albeit with a higher average computational time of 28,0213 seconds. Meanwhile, PSO remained more efficient with a time of 25,4761 seconds, while still delivering competitive results (DBI of 0,5878 and Silhouette Score of 0,5559). Overall, PSO demonstrated superior capability in identifying optimal cluster centroids quickly and accurately compared to GA and FPA. Consequently, the PSO-optimized K-Prototypes model was selected as the final model and implemented in a Streamlit-based application to facilitate automated clustering and support data-driven decision-making.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Hindrayani, Kartika Maulida	NIDN0009099205	kartika.maulida.ds@upnjatim.ac.id
Thesis advisor	Saputra, Wahyu Syaifullah Jauharis	NIDN0725088601	wahyu.s.j.saputra.if@upnjatim.ac.id

Subjects:

Q Science > QA Mathematics > QA76.6 Computer Programming

Divisions:

Faculty of Computer Science > Departemen of Data Science

Depositing User:

Lisya Septyo Ningrum

Date Deposited:

19 Jun 2025 01:49

Last Modified:

19 Jun 2025 01:49

URI:

https://repository.upnjatim.ac.id/id/eprint/38266