| 1 | Second-Dimension Spatial Association (SDA) | SDA extracts the geographical and environmental information outside sample locations for spatial prediction. SDA improves prediction accuracy compared with Random Forest Kriging, one of the best-performed spatial machine learning models. | Song et al., 2022, JAG |
| 2 | Optimal Parameters-based Geographical Detector (OPGD) | OPGD is used to characterising spatial heterogeneity, identifying geographical factors and interactive impacts of factors, and estimating risks. Software: R package “GD”. | Song et al., 2020, GRS |
| 3 | Interactive Detector for Spatial Associations (IDSA) | IDSA is used to estimate power of interactive determinants (PID) from a spatial perspective. The IDSA model considers spatial heterogeneity, spatial autocorrelation, and spatial fuzzy overlay of multiple explanatory variables for calculating PID. Software: R package “IDSA”. | Song et al., 2021, IJGIS |
| 4 | Spatial Trade-Off Relation (STOR) | STOR is used to assess quantity-quality trade-off relations of sustainability indicators (of infrastructure) from a spatial perspective. | Song et al., 2021, JAG |
| 5 | MFSD: Spatial Decision Making | Model-driven fuzzy spatial multi-criteria decision making (MFSD) approach is used to generate data and model driven sustainable road infrastructure performance indicators and reduce potential biases of human decisions. | Song et al., 2021, RSER |
| 6 | Spatial Heterogeneity-based Segmentation (SHS) | SHS is used to segment spatial line data. The SHS model is implemented in redefining road segments with high-resolution sensor monitoring data. Software: R package “HS”. | Song et al., 2020, IEEE-ITS |
| 7 | Line Segment Regression Kriging (SRK) | SRK is used for spatial interpolation and prediction of line segment data, such as road and traffic data. Kriging covariance functions is estimated based on the covariance between any two line segments. | Song et al., 2019, IEEE-ITS (ESI Highly Citied Paper) |
| 8 | D2SFCA: Spatiotemporal Accessibility | Travel time-based modified kernel density two-step floating catchment area (MKD2SFCA) model is used to compute population accessibility to public facilities, such as hospitals. | Song et al., 2018, GRS (ESI Highly Cited Paper) |
| 9 | EOSI: Earth Data for Sustainable Infrastructure | Earth observation has great potentials for sustainable infrastructure development. EOSI benefits about 85% of infrastructure influenced SDGs and 61% of all 169 SDG targets, but Earth observation is only implemented in 15% of infrastructure influenced SDG targets, and 70% of infrastructure influenced targets that can be directly or indirectly derived from Earth observation data have not been included in current SDG indicators. | Song et al., 2021, RS |
| 10 | Spatial Big Data Method for Redefining Cities | Nature cities or central urban regions can be identified using a spatial big data-driven method. The commonly used data include point of interests (POI) and social media data. | Song et al., 2018 IJGIS |
| 11 | Geographically Optimal Zones-based Heterogeneity (GOZH) | GOZH is used to identify individual and interactive determinants of geographical attributes (e.g., global soil moisture) across a large study area. | Luo, P., Song, Y.*, et al., 2022, ISPRS P&RS |
| 12 | Wavelet Geographically Weighted Regression (WGWR) | WGWR is used for the spatial prediction of soil properties with spectroscopic data. | Song et al., 2021, SR |
| 13 | Robust Geographical Detector (RGD) | RGD provides robust estimations of the power of determinants of spatial explanatory variables. | Zhang, Song*, et al, 2022, JAG |
Dr. Yongze Song's Homepage 