Open spatial sciences: an introduction

Páez, Antonio

doi:10.1007/s10109-021-00364-4

Cited by 19 publications

(11 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For ML models, R packages ranger for random forests (Wright and Ziegler, 2017), nnet for ANN models (Venables and Ripley, 2002), and e1071 (https://cran.r-project.org/web/packages/e1071/index.html, accessed August 30, 2021) for SVM models, were used. For promoting the reproducible research practice in quantitative geography and spatial data science (Brunsdon and Comber, 2021; Kedron et al, 2021; Páez, 2021), we maintained our analysis as reproducible as possible. Analysis codes and data, which reproduce our research results in machines at any platform that can run R, are shared on the authors' webpage (https://github.com/biogeokim/ml_sac_prediction).…”

Section: Methodsmentioning

confidence: 99%

Three Common Machine Learning Algorithms Neither Enhance Prediction Accuracy Nor Reduce Spatial Autocorrelation in Residuals: An Analysis of Twenty‐five Socioeconomic Data Sets

Song

Kim

2022

Geographical Analysis

View full text Add to dashboard Cite

Machine learning (ML) is being applied in an increasing volume of geographical research. However, the aspects of spatial autocorrelation (SAC) in the residuals produced by ML models have been understudied compared to the benefit of ML, namely, reduction of prediction errors. In this study, we examined the relationship between predictive accuracy and the reduction in the residual SAC for 597 variables from 25 geographical socio‐economic data sets using spatial and nonspatial cross‐validation of three ML algorithms such as random forests, support vector machine, and artificial neural network (ANN) to provide an extensive empirical diagnosis—but not a definitive theory—of the relationship between SAC and ML. Our results highlighted that the ML algorithms with tuned hyperparameters yielded marginal predictive accuracy gains and the minimal decreases in residual SAC. ANN revealed lower accuracy and higher reduction in the residual SAC than others. This implies ML algorithms in geographical research in socio‐economic domains would not always result in higher prediction accuracy. We suggest that ML in geographical research should be cautiously employed when the main objective is related to the residual SAC. We also showed that spatial cross‐validation neither improves predictive accuracy substantially nor reduce the residual SAC effectively.

show abstract

Section: Methodsmentioning

confidence: 99%

Three Common Machine Learning Algorithms Neither Enhance Prediction Accuracy Nor Reduce Spatial Autocorrelation in Residuals: An Analysis of Twenty‐five Socioeconomic Data Sets

Song

Kim

2022

Geographical Analysis

View full text Add to dashboard Cite

show abstract

“…Open science developed in response to calls within the scientific community for greater transparency and reproducibility (Morin et al, 2012; Rocchini & Neteler, 2012; Stodden et al, 2013; Watson, 2015). These calls for transparency and reproducibility in the sciences are not new; the advent of the scientific journal was, after all, in response to those needs (Páez, 2021). However, centuries later, science still struggles to accomplish this.…”

Section: Introductionmentioning

confidence: 99%

Integrating GRASS GIS and Jupyter Notebooks to facilitate advanced geospatial modeling education

Haedrich

Petras

Petrášová

et al. 2023

Transactions in GIS

View full text Add to dashboard Cite

Open education materials are critical for the advancement of open science and the development of open‐source software. These accessible and transparent materials provide an important pathway for sharing both standard geospatial analysis workflows and advanced research methods. Computational notebooks allow users to share live code with in‐line visualizations and narrative text, making them a powerful interactive teaching tool for geospatial analytics. Specifically, Jupyter Notebooks are quickly becoming a standard format in open education. In this article, we introduce a new GRASS GIS package, grass.jupyter, that enhances the existing GRASS Python API to allow Jupyter Notebook users to easily manage and visualize GRASS data including spatiotemporal datasets. While there are many Python‐based geospatial libraries available for use in Jupyter Notebooks, GRASS GIS has extensive geospatial functionality including support for multi‐temporal analysis and dynamic simulations, making it a powerful teaching tool for advanced geospatial analytics. We discuss the development of grass.jupyter and demonstrate how the package facilitates teaching open‐source geospatial modeling with a collection of Jupyter Notebooks designed for a graduate‐level geospatial modeling course. The open education notebooks feature spatiotemporal data visualizations, hydrologic modeling, and spread simulations such as the spread of invasive species and urban growth.

show abstract

“…Discussion is supported by the use of the small synthetic example of Shen (1998) and empirical data drawn from the 2016 Transportation Tomorrow Survey of the Greater Toronto and Hamilton Area in Ontario, Canada. In the spirit of openness of research in the spatial sciences (Brunsdon and Comber, 2021;Páez, 2021) this paper has a companion open data product (Arribas-Bel et al, 2021), and all code is available for replicability and reproducibility purposes at https://github.com/soukhova/Spatial-Availability-Measure.…”

Section: Introductionmentioning

confidence: 99%

Introducing spatial availability, a singly-constrained measure of competitive accessibility

Soukhov¹,

Páez²,

Higgins³

et al. 2022

Preprint

View full text Add to dashboard Cite

Accessibility indicators are widely used in transportation, urban, and healthcare planning, among many other applications. These measures are weighted sums of reachable opportunities from a given origin conditional on the cost of movement, and are estimates of the potential for spatial interaction. Over time, various proposals have been forwarded to improve their interpretability, mainly by introducing competition. In this paper, we demonstrate how a widely used measure of accessibility with congestion fails to properly match the opportunity-seeking population. We then propose an alternative formulation of accessibility with competition, a measure we call _spatial availability_. This measure results from using balancing factors that are equivalent to imposing a single constraint on conventional gravity-based accessibility. Further, we demonstrate how Two-Stage Floating Catchment Area (2SFCA) methods can be reconceptualized as singly-constrained accessibility. To illustrate the application of spatial availability and compare it to other relevant measures, we use data from the 2016 Transportation Tomorrow Survey of the Greater Golden Horseshoe area in southern Ontario, Canada.

show abstract

Open spatial sciences: an introduction

Cited by 19 publications

References 49 publications

Three Common Machine Learning Algorithms Neither Enhance Prediction Accuracy Nor Reduce Spatial Autocorrelation in Residuals: An Analysis of Twenty‐five Socioeconomic Data Sets

Three Common Machine Learning Algorithms Neither Enhance Prediction Accuracy Nor Reduce Spatial Autocorrelation in Residuals: An Analysis of Twenty‐five Socioeconomic Data Sets

Integrating GRASS GIS and Jupyter Notebooks to facilitate advanced geospatial modeling education

Introducing spatial availability, a singly-constrained measure of competitive accessibility

Contact Info

Product

Resources

About