Combining supervised and unsupervised methods in tourism visitor data

Weksi Budiaji and Vebriana and Juwarin Pancawati (2022) Combining supervised and unsupervised methods in tourism visitor data. Journal of Information Technology and Its Utilization, 5 (1): 3. pp. 14-17. ISSN 2654-802X

[thumbnail of Jurnal_Weksi Budiaji_Universitas Sultan Ageng Tirtayasa_2022-3.pdf]
Jurnal_Weksi Budiaji_Universitas Sultan Ageng Tirtayasa_2022-3.pdf - Published Version
Available under License Creative Commons Attribution Non-commercial Share Alike.

Download (226kB) | Preview


Combining supervised and unsupervisedmethod can assist in the data analysis process. This research aims to apply a supervised method, i.e. Poisson regression, that is followed by an unsupervised method, namely cluster analysis of the visitors in a tourism dataset.The samples were taken 80 persons purposively from the visitors of the Flower Garden X in Serang Regency, Banten Province. The dataset consists of the number of visits, travel cost, income/ stipend per month, gender, age, distance from the place of origin, and perception, which is formed by 11 questions of facilities and services. The Poisson regression was applied in the 30, 40, and 50 bootstrap samples resulted in the perception as the significant features. Then, medoid-based cluster analysis, i.e. pam and simple k-medoids, in the perception dataset was applied. They compared simple matching andcooccurrence distances and were validated via medoidbased shadow value. It grouped the visitors into fiveclusters as the most suitable number of clusters. The combined methods of supervised and unsupervisedprovided the cleanliness as the important indicator. The improvement of the tourism object had to be focus on the cleanliness aspect.

Item Type: Article
Uncontrolled Keywords: Bootstrap samples; Cluster analysis; Poisson regression; Supervised method; Unsupervised method.
Subjects: Urban & Regional Technology & Development > Recreation
Economics and Business
Economics and Business > Domestic Commerce, Marketing, & Economics
Depositing User: Djaenudin djae Mohamad
Date Deposited: 07 Mar 2023 12:17
Last Modified: 13 Mar 2023 03:42

Actions (login required)

View Item
View Item