XGBOOST MODEL FOR PARK VISITATION PREDICTION IN A MID-SIZE CITY

No Thumbnail Available

Date

2023-04-05

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

Parks have a significant impact on residents’ health and social activities. By using smartphone mobility data tracking the activities of 28 parks in the College Station and Bryan Metropolitan area of Texas, USA, I present the temporal and spatial patterns of park usage within a two-year timeframe. I model the effects of the socio-economic, built environment, climate, surrounding points of interest (POI), and spatial/accessibility factors on park visitations through a machine learning model. The results show that climate change and nearby POIs such as restaurants and gas stations are significant factors enhancing park visitations while having hotels and apartment complexes are not. The study also reveals how smartphone mobility data can be applied to case studies investigating urban design/planning and understanding the social and adjacent points of interest associated with urban greenspaces. It provides empirical evidence on park visitations as well as what factors future planners, landscape architects, and park managers should consider when deciding on park investment and planning decisions for mid-sized cities.

Description

Keywords

human mobility, park visitation, xgboost

Citation