ISPRS International Journal of Geo-Information

Research

9 pages, 1302 KiB

Open AccessCommunication

A Forest of Forests: A Spatially Weighted and Computationally Efficient Formulation of Geographical Random Forests

by Stefanos Georganos and Stamatis Kalogirou

ISPRS Int. J. Geo-Inf. 2022, 11(9), 471; https://doi.org/10.3390/ijgi11090471 - 31 Aug 2022

Cited by 17 | Viewed by 5760

The aim of this paper is to present developments of an advanced geospatial analytics algorithm that improves the prediction power of a random forest regression model while addressing the issue of spatial dependence commonly found in geographical data. We applied the methodology to [...] Read more.

The aim of this paper is to present developments of an advanced geospatial analytics algorithm that improves the prediction power of a random forest regression model while addressing the issue of spatial dependence commonly found in geographical data. We applied the methodology to a simple model of mean household income in the European Union regions to allow easy understanding and reproducibility of the analysis. The results are encouraging and suggest an improvement in the prediction power compared to previous techniques. The algorithm has been implemented in R and is available in the updated version of the SpatialML package in the CRAN repository. Full article

(This article belongs to the Special Issue Advances in Machine Learning and Statistical Analysis of Geographical Data)

► Show Figures

Figure 1

21 pages, 3465 KiB

Open AccessArticle

Incorporating Spatial Autocorrelation in Machine Learning Models Using Spatial Lag and Eigenvector Spatial Filtering Features

by Xiaojian Liu, Ourania Kounadi and Raul Zurita-Milla

ISPRS Int. J. Geo-Inf. 2022, 11(4), 242; https://doi.org/10.3390/ijgi11040242 - 07 Apr 2022

Cited by 12 | Viewed by 5720

Abstract

Applications of machine-learning-based approaches in the geosciences have witnessed a substantial increase over the past few years. Here we present an approach that accounts for spatial autocorrelation by introducing spatial features to the models. In particular, we explore two types of spatial features, [...] Read more.

Applications of machine-learning-based approaches in the geosciences have witnessed a substantial increase over the past few years. Here we present an approach that accounts for spatial autocorrelation by introducing spatial features to the models. In particular, we explore two types of spatial features, namely spatial lag and eigenvector spatial filtering (ESF). These features are used within the widely used random forest (RF) method, and their effect is illustrated on two public datasets of varying sizes (Meuse and California housing datasets). The least absolute shrinkage and selection operator (LASSO) is used to determine the best subset of spatial features, and nested cross-validation is used for hyper-parameter tuning and performance evaluation. We utilize Moran’s I and local indicators of spatial association (LISA) to assess how spatial autocorrelation is captured at both global and local scales. Our results show that RF models combined with either spatial lag or ESF features yield lower errors (up to 33% different) and reduce the global spatial autocorrelation of the residuals (up to 95% decrease in Moran’s I) compared to the RF model with no spatial features. The local autocorrelation patterns of the residuals are weakened as well. Compared to benchmark geographically weighted regression (GWR) models, the RF models with spatial features yielded more accurate models with similar levels of global and local autocorrelation in the prediction residuals. This study reveals the effectiveness of spatial features in capturing spatial autocorrelation and provides a generic machine-learning modelling workflow for spatial prediction. Full article

(This article belongs to the Special Issue Advances in Machine Learning and Statistical Analysis of Geographical Data)

► Show Figures

Figure 1

19 pages, 4533 KiB

Open AccessArticle

Domain Adaptation for Semantic Segmentation of Historical Panchromatic Orthomosaics in Central Africa

by Nicholus Mboga, Stefano D’Aronco, Tais Grippa, Charlotte Pelletier, Stefanos Georganos, Sabine Vanhuysse, Eléonore Wolff, Benoît Smets, Olivier Dewitte, Moritz Lennert and Jan Dirk Wegner

ISPRS Int. J. Geo-Inf. 2021, 10(8), 523; https://doi.org/10.3390/ijgi10080523 - 01 Aug 2021

Cited by 6 | Viewed by 2902

Abstract

Multitemporal environmental and urban studies are essential to guide policy making to ultimately improve human wellbeing in the Global South. Land-cover products derived from historical aerial orthomosaics acquired decades ago can provide important evidence to inform long-term studies. To reduce the manual labelling [...] Read more.

Multitemporal environmental and urban studies are essential to guide policy making to ultimately improve human wellbeing in the Global South. Land-cover products derived from historical aerial orthomosaics acquired decades ago can provide important evidence to inform long-term studies. To reduce the manual labelling effort by human experts and to scale to large, meaningful regions, we investigate in this study how domain adaptation techniques and deep learning can help to efficiently map land cover in Central Africa. We propose and evaluate a methodology that is based on unsupervised adaptation to reduce the cost of generating reference data for several cities and across different dates. We present the first application of domain adaptation based on fully convolutional networks for semantic segmentation of a dataset of historical panchromatic orthomosaics for land-cover generation for two focus cities Goma-Gisenyi and Bukavu. Our experimental evaluation shows that the domain adaptation methods can reach an overall accuracy between 60% and 70% for different regions. If we add a small amount of labelled data from the target domain, too, further performance gains can be achieved. Full article

(This article belongs to the Special Issue Advances in Machine Learning and Statistical Analysis of Geographical Data)

► Show Figures

Figure 1

15 pages, 3561 KiB

Open AccessArticle

Could Historical Mortality Data Predict Mortality Due to Unexpected Events?

by Panagiotis Andreopoulos, Kleomenis Kalogeropoulos, Alexandra Tragaki and Nikolaos Stathopoulos

ISPRS Int. J. Geo-Inf. 2021, 10(5), 283; https://doi.org/10.3390/ijgi10050283 - 29 Apr 2021

Cited by 2 | Viewed by 2045

Abstract

Research efforts focused on developing a better understanding of the evolution of mortality over time are considered to be of significant interest—not just to the demographers. Mortality can be expressed with different parameters through multiparametric prediction models. Based on the Beta Gompertz generalized [...] Read more.

Research efforts focused on developing a better understanding of the evolution of mortality over time are considered to be of significant interest—not just to the demographers. Mortality can be expressed with different parameters through multiparametric prediction models. Based on the Beta Gompertz generalized Makeham (BGGM) distribution, this study aims to evaluate and map four of such parameters for 22 countries of the European Union, over the period 1960–2045. The BGGM probabilistic distribution is a multidimensional model, which can predict using the corresponding probabilistic distribution with the following parameters: infant mortality (parameter θ), population aging (parameter ξ), and individual and population mortality due to unexpected exogenous factors/events (parameters κ and λ, respectively). This work focuses on the random risk factor (λ) that can affect the entire population, regardless of age and gender, with increasing mortality depicting developments and trends, both temporally (past–present–future) and spatially (22 countries). Moreover, this study could help policymakers in the field of health provide solutions in terms of mortality. Mathematical models like BGGM can be used to achieve and highlight probable cyclical repetitions of sudden events (such as Covid-19) in different time series for different geographical areas. GIS context is used to map the spatial patterns of this estimated parameter as well as these variations during the examined period for both men and women. Full article

(This article belongs to the Special Issue Advances in Machine Learning and Statistical Analysis of Geographical Data)

► Show Figures

Figure 1

17 pages, 3973 KiB

Open AccessArticle

Non-Local Feature Search Network for Building and Road Segmentation of Remote Sensing Image

by Cheng Ding, Liguo Weng, Min Xia and Haifeng Lin

ISPRS Int. J. Geo-Inf. 2021, 10(4), 245; https://doi.org/10.3390/ijgi10040245 - 07 Apr 2021

Cited by 15 | Viewed by 3282

Abstract

Building and road extraction from remote sensing images is of great significance to urban planning. At present, most of building and road extraction models adopt deep learning semantic segmentation method. However, the existing semantic segmentation methods did not pay enough attention to the [...] Read more.

Building and road extraction from remote sensing images is of great significance to urban planning. At present, most of building and road extraction models adopt deep learning semantic segmentation method. However, the existing semantic segmentation methods did not pay enough attention to the feature information between hidden layers, which led to the neglect of the category of context pixels in pixel classification, resulting in these two problems of large-scale misjudgment of buildings and disconnection of road extraction. In order to solve these problem, this paper proposes a Non-Local Feature Search Network (NFSNet) that can improve the segmentation accuracy of remote sensing images of buildings and roads, and to help achieve accurate urban planning. By strengthening the exploration of hidden layer feature information, it can effectively reduce the large area misclassification of buildings and road disconnection in the process of segmentation. Firstly, a Self-Attention Feature Transfer (SAFT) module is proposed, which searches the importance of hidden layer on channel dimension, it can obtain the correlation between channels. Secondly, the Global Feature Refinement (GFR) module is introduced to integrate the features extracted from the backbone network and SAFT module, it enhances the semantic information of the feature map and obtains more detailed segmentation output. The comparative experiments demonstrate that the proposed method outperforms state-of-the-art methods, and the model complexity is the lowest. Full article

(This article belongs to the Special Issue Advances in Machine Learning and Statistical Analysis of Geographical Data)

► Show Figures

Figure 1

23 pages, 6357 KiB

Open AccessArticle

Machine Learning-Based Processing Proof-of-Concept Pipeline for Semi-Automatic Sentinel-2 Imagery Download, Cloudiness Filtering, Classifications, and Updates of Open Land Use/Land Cover Datasets

by Tomáš Řezník, Jan Chytrý and Kateřina Trojanová

ISPRS Int. J. Geo-Inf. 2021, 10(2), 102; https://doi.org/10.3390/ijgi10020102 - 23 Feb 2021

Cited by 6 | Viewed by 3386

Abstract

Land use and land cover are continuously changing in today’s world. Both domains, therefore, have to rely on updates of external information sources from which the relevant land use/land cover (classification) is extracted. Satellite images are frequent candidates due to their temporal and [...] Read more.

Land use and land cover are continuously changing in today’s world. Both domains, therefore, have to rely on updates of external information sources from which the relevant land use/land cover (classification) is extracted. Satellite images are frequent candidates due to their temporal and spatial resolution. On the contrary, the extraction of relevant land use/land cover information is demanding in terms of knowledge base and time. The presented approach offers a proof-of-concept machine-learning pipeline that takes care of the entire complex process in the following manner. The relevant Sentinel-2 images are obtained through the pipeline. Later, cloud masking is performed, including the linear interpolation of merged-feature time frames. Subsequently, four-dimensional arrays are created with all potential training data to become a basis for estimators from the scikit-learn library; the LightGBM estimator is then used. Finally, the classified content is applied to the open land use and open land cover databases. The verification of the provided experiment was conducted against detailed cadastral data, to which Shannon’s entropy was applied since the number of cadaster information classes was naturally consistent. The experiment showed a good overall accuracy (OA) of 85.9%. It yielded a classified land use/land cover map of the study area consisting of 7188 km2 in the southern part of the South Moravian Region in the Czech Republic. The developed proof-of-concept machine-learning pipeline is replicable to any other area of interest so far as the requirements for input data are met. Full article

(This article belongs to the Special Issue Advances in Machine Learning and Statistical Analysis of Geographical Data)

► Show Figures

Figure 1

14 pages, 6433 KiB

Open AccessArticle

Multi-View Instance Matching with Learned Geometric Soft-Constraints

by Ahmed Samy Nassar, Sébastien Lefèvre and Jan Dirk Wegner

ISPRS Int. J. Geo-Inf. 2020, 9(11), 687; https://doi.org/10.3390/ijgi9110687 - 18 Nov 2020

Cited by 3 | Viewed by 2323

Abstract

We present a new approach for matching urban object instances across multiple ground-level images for the ultimate goal of city-scale mapping of objects with high positioning accuracy. What makes this task challenging is the strong change in view-point, different lighting conditions, high similarity [...] Read more.

We present a new approach for matching urban object instances across multiple ground-level images for the ultimate goal of city-scale mapping of objects with high positioning accuracy. What makes this task challenging is the strong change in view-point, different lighting conditions, high similarity of neighboring objects, and variability in scale. We propose to turn object instance matching into a learning task, where image-appearance and geometric relationships between views fruitfully interact. Our approach constructs a Siamese convolutional neural network that learns to match two views of the same object given many candidate image cut-outs. In addition to image features, we propose utilizing location information about the camera and the object to support image evidence via soft geometric constraints. Our method is compared to existing patch matching methods to prove its edge over state-of-the-art. This takes us one step closer to the ultimate goal of city-wide object mapping from street-level imagery to benefit city administration. Full article

(This article belongs to the Special Issue Advances in Machine Learning and Statistical Analysis of Geographical Data)

► Show Figures

Figure 1

23 pages, 16541 KiB

Open AccessArticle

Rainfall-Induced Shallow Landslide Susceptibility Mapping at Two Adjacent Catchments Using Advanced Machine Learning Algorithms

by Ananta Man Singh Pradhan and Yun-Tae Kim

ISPRS Int. J. Geo-Inf. 2020, 9(10), 569; https://doi.org/10.3390/ijgi9100569 - 29 Sep 2020

Cited by 37 | Viewed by 3915

Abstract

Landslides impact on human activities and socio-economic development, especially in mountainous areas. This study focuses on the comparison of the prediction capability of advanced machine learning techniques for the rainfall-induced shallow landslide susceptibility of Deokjeokri catchment and Karisanri catchment in South Korea. The [...] Read more.

Landslides impact on human activities and socio-economic development, especially in mountainous areas. This study focuses on the comparison of the prediction capability of advanced machine learning techniques for the rainfall-induced shallow landslide susceptibility of Deokjeokri catchment and Karisanri catchment in South Korea. The influencing factors for landslides, i.e., topographic, hydrologic, soil, forest, and geologic factors, are prepared from various sources based on availability, and a multicollinearity test is also performed to select relevant causative factors. The landslide inventory maps of both catchments are obtained from historical information, aerial photographs and performed field surveys. In this study, Deokjeokri catchment is considered as a training area and Karisanri catchment as a testing area. The landslide inventories contain 748 landslide points in training and 219 points in testing areas. Three landslide susceptibility maps using machine learning models, i.e., Random Forest (RF), Extreme Gradient Boosting (XGBoost) and Deep Neural Network (DNN), are prepared and compared. The outcomes of the analyses are validated using the landslide inventory data. A receiver operating characteristic curve (ROC) method is used to verify the results of the models. The results of this study show that the training accuracy of RF is 0.756 and the testing accuracy is 0.703. Similarly, the training accuracy of XGBoost is 0.757 and testing accuracy is 0.74. The prediction of DNN revealed acceptable agreement between the susceptibility map and the existing landslides, with a training accuracy of 0.855 and testing accuracy of 0.802. The results showed that the DNN model achieved lower prediction error and higher accuracy results than other models for shallow landslide modeling in the study area. Full article

(This article belongs to the Special Issue Advances in Machine Learning and Statistical Analysis of Geographical Data)

► Show Figures

Graphical abstract

Journal Menu

Journal Browser

Advances in Machine Learning and Statistical Analysis of Geographical Data

Share This Special Issue

Special Issue Editors

Special Issue Information

Keywords

Published Papers (8 papers)

Research

Further Information

Guidelines

MDPI Initiatives

Follow MDPI