GNNWR: an open-source package of spatiotemporal intelligent regression methods for modeling spatial and temporal nonstationarity DOI Creative Commons
Ziyu Yin, Jiale Ding, Y.Y. Liu

et al.

Geoscientific model development, Journal Year: 2024, Volume and Issue: 17(22), P. 8455 - 8468

Published: Nov. 28, 2024

Abstract. Spatiotemporal regression is a crucial method in geography for discerning spatiotemporal nonstationarity geographical relationships and has found widespread application across diverse research domains. This study implements two innovative intelligent models, i.e., Geographically Neural Network Weighted Regression (GNNWR) Temporally (GTNNWR), which use neural networks to estimate nonstationarity. Due the higher accuracy generalization ability, these models have been widely used various fields of scientific research. To facilitate GNNWR GTNNWR addressing nonstationary processes, Python-based package developed. article details implementation introduces package, enabling users efficiently apply cutting-edge techniques. Validation conducted through case studies. The first involves verification using air quality data from China, while second employs offshore dissolved silicate concentration Zhejiang Province validate GTNNWR. results studies underscore effectiveness yielding outcomes notable accuracy. contribution anticipates significant role developed supporting future that will leverage big

Language: Английский

SGWR: similarity and geographically weighted regression DOI Creative Commons
M. Naser Lessani, Zhenlong Li

International Journal of Geographical Information Science, Journal Year: 2024, Volume and Issue: 38(7), P. 1232 - 1255

Published: April 17, 2024

Geographically weighted regression (GWR) offers a local approach to modeling spatial data, considering geographical location and relationships between observations. A salient feature of GWR is the emphasis on proximity, in accordance with Tobler's First Law Geography, which assumes that closer entities have greater influence target location. Traditional models been augmented consider various forms physical distances aimed at enhancing model performance, they often disregarded potential other data attributes, shortcoming extends most extensions. In this study, we introduce novel weight matrix construction, integrates attribute similarity alongside conventional geographically matrix. The two weights are integrated manner results improved performance. proposed model, called Similarity Weighted Regression or SGWR, was applied five distinct datasets: housing prices, crime rates, three health outcomes including mental health, depression, HIV. Results show SGWR significantly performance based several statistical measures, outperforming global traditional GWR.

Language: Английский

Citations

8
Research on regional economic development and natural disaster risk assessment under the goal of carbon peak and carbon neutrality: A case study in Chengdu-Chongqing economic circle DOI
Xin Zhang, Hao Luo, Xiaoyu Zeng

et al.

Land Use Policy, Journal Year: 2024, Volume and Issue: 143, P. 107206 - 107206

Published: May 28, 2024

Language: Английский

Citations

7
The relationship between accessibility and land prices: A focus on accessibility to transit in the 15-min city DOI
Zijuan Yin, Wenquan Li, Congcong Li

et al.

Travel Behaviour and Society, Journal Year: 2024, Volume and Issue: 38, P. 100914 - 100914

Published: Sept. 28, 2024

Language: Английский

Citations

4
Using an attention-based architecture to incorporate context similarity into spatial non-stationarity estimation DOI
Sensen Wu, Jiale Ding, Ruoxu Wang

et al.

International Journal of Geographical Information Science, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 24

Published: Feb. 10, 2025

Language: Английский

Citations

0
Deciphering global marine product trade dynamics: Patterns, drivers, and policy insights DOI
Yu Sun,

Feng Lian,

Zhongzhen Yang

et al.

Marine Policy, Journal Year: 2025, Volume and Issue: 177, P. 106681 - 106681

Published: March 17, 2025

Language: Английский

Citations

0
Uncovering differences in the spatial structure of intercity interactive networks described by multi-source migration flow: From the multi-hierarchical perspective DOI
Shimei Wei, Jinghu Pan

Journal of Geographical Sciences, Journal Year: 2025, Volume and Issue: 35(5), P. 1049 - 1079

Published: May 1, 2025

Language: Английский

Citations

0
Learning dynamic relational heterogeneity for spatiotemporal prediction with geographical meta-knowledge DOI
Kaiqi Chen, Xiaoyong Tan, Min Deng

et al.

International Journal of Geographical Information Science, Journal Year: 2025, Volume and Issue: unknown, P. 1 - 30

Published: May 20, 2025

Language: Английский

Citations

0
A neural network model to optimize the measure of spatial proximity in geographically weighted regression approach: a case study on house price in Wuhan DOI
Jiale Ding,

Wenying Cen,

Sensen Wu

et al.

International Journal of Geographical Information Science, Journal Year: 2024, Volume and Issue: 38(7), P. 1315 - 1335

Published: April 25, 2024

The estimation of spatial heterogeneity within real estate markets holds significant importance in house price modelling. However, employing a single or straightforward distance to measure proximity is probably insufficient complex urban areas, thereby resulting an inadequate modelling heterogeneity. To address this issue, paper incorporates multiple measures neural network framework achieve optimized (OSP). Consequently, geographically weighted regression model with (osp-GNNWR) devised for the purpose spatially heterogeneous modeling. Trained as unified model, osp-GNNWR obviates need separate pretraining OSP. This enables OSP delineate modeled process through post hoc calculated value. Through simulation experiments and real-world case study on prices, proposed reaches more accurate descriptions diverse processes exhibits better overall performance. interpretable results Wuhan demonstrate efficacy addressing markets, suggesting its potential predicting geographical phenomena.

Language: Английский

Citations

3
Study on the impact of the high-speed rail network on industrial structure upgrading DOI

Qifen Zha,

Zhen Liu, Jian Wang

et al.

Research in Transportation Business & Management, Journal Year: 2023, Volume and Issue: 51, P. 101044 - 101044

Published: Oct. 10, 2023

Language: Английский

Citations

8
GNNWR: An Open-Source Package of Spatiotemporal Intelligent Regression Methods for Modeling Spatial and Temporal Non-Stationarity DOI Creative Commons
Ziyu Yin, Jiale Ding, Y.Y. Liu

et al.

Published: May 29, 2024

Abstract. Spatiotemporal regression is a crucial method in geography for discerning spatiotemporal non-stationarity geographical relationships, which has found widespread application across diverse research domains. This study implements two innovative intelligent models, namely geographically neural network weighted (GNNWR) and temporally (GTNNWR), integrating the framework networks. Demonstrating superior accuracy generalization capabilities large-scale data environments compared to traditional methods, these models have emerged as prominent tools. To facilitate seamless of GNNWR GTNNWR addressing non-stationary processes, Python-based package, GNNWR, been developed. article details implementation introduces enabling users efficiently apply cutting-edge techniques. Validation package conducted through case studies. The first involves verification using air quality from China, while second employs offshore dissolved silicate concentration Zhejiang Province validate GTNNWR. results studies underscore effectiveness yielding outcomes notable accuracy. contribution anticipates significant role developed supporting future that leverages big

Language: Английский

Citations

1