Matches in SemOpenAlex for { <https://semopenalex.org/work/W4378647911> ?p ?o ?g. }
- W4378647911 endingPage "103364" @default.
- W4378647911 startingPage "103364" @default.
- W4378647911 abstract "Random cross-validation (CV) is often used to evaluate geospatial machine learning models, particularly when a limited amount of sample data are available, and collecting an extra test set is unfeasible. However, the prediction locations can be substantially different from the available sample, leading to over-optimistic evaluation results. This has fostered the development of spatial CV methods. Yet these methods only focus on spatial autocorrelation and cannot sufficiently guarantee that the validation subset is a good proxy of the test set with significant differences. In this paper, we propose the spatial+ cross-validation (SP-CV) method. This method, which considers both the geographic and feature spaces, is composed of two stages. The first stage addresses spatial autocorrelation issues by using agglomerative hierarchical clustering to divide the available sample into blocks. The second stage deals with multiple sources of differences. It uses cluster ensembles to split the blocks into training and validation folds based on the locations of the sample data and the values of the covariates and target variable. The proposed method is compared against random and block CV methods in a series of experiments with Amazon basin above ground biomass and California houseprice datasets. Our results show that SP-CV provided the smallest error differences with respect to the reference error. This means that SP-CV produced more representative splits and led to more reliable model evaluations. It suggests that a reliable model evaluation requires to consider both the geographic and the feature spaces in a comprehensive manner." @default.
- W4378647911 created "2023-05-30" @default.
- W4378647911 creator A5055884274 @default.
- W4378647911 creator A5075715085 @default.
- W4378647911 creator A5075724211 @default.
- W4378647911 date "2023-07-01" @default.
- W4378647911 modified "2023-09-26" @default.
- W4378647911 title "Spatial+: A new cross-validation method to evaluate geospatial machine learning models" @default.
- W4378647911 cites W1989179819 @default.
- W4378647911 cites W2023320048 @default.
- W4378647911 cites W2040615655 @default.
- W4378647911 cites W2050179592 @default.
- W4378647911 cites W2057779644 @default.
- W4378647911 cites W2105897946 @default.
- W4378647911 cites W2122447387 @default.
- W4378647911 cites W2130956715 @default.
- W4378647911 cites W2141807666 @default.
- W4378647911 cites W2149230623 @default.
- W4378647911 cites W2155544089 @default.
- W4378647911 cites W2407205744 @default.
- W4378647911 cites W2560136348 @default.
- W4378647911 cites W2605018929 @default.
- W4378647911 cites W2729033468 @default.
- W4378647911 cites W2790063226 @default.
- W4378647911 cites W2793997912 @default.
- W4378647911 cites W2794916302 @default.
- W4378647911 cites W2803051956 @default.
- W4378647911 cites W2819013944 @default.
- W4378647911 cites W2896482460 @default.
- W4378647911 cites W2897898131 @default.
- W4378647911 cites W2898407312 @default.
- W4378647911 cites W2900086371 @default.
- W4378647911 cites W2907788294 @default.
- W4378647911 cites W2910729503 @default.
- W4378647911 cites W2911964244 @default.
- W4378647911 cites W2952516441 @default.
- W4378647911 cites W2972629016 @default.
- W4378647911 cites W3014395405 @default.
- W4378647911 cites W3030581860 @default.
- W4378647911 cites W3036793583 @default.
- W4378647911 cites W3049094422 @default.
- W4378647911 cites W3085784695 @default.
- W4378647911 cites W3136230353 @default.
- W4378647911 cites W3169855463 @default.
- W4378647911 cites W3194508523 @default.
- W4378647911 cites W4220932724 @default.
- W4378647911 cites W4224247190 @default.
- W4378647911 cites W4229048457 @default.
- W4378647911 doi "https://doi.org/10.1016/j.jag.2023.103364" @default.
- W4378647911 hasPublicationYear "2023" @default.
- W4378647911 type Work @default.
- W4378647911 citedByCount "0" @default.
- W4378647911 crossrefType "journal-article" @default.
- W4378647911 hasAuthorship W4378647911A5055884274 @default.
- W4378647911 hasAuthorship W4378647911A5075715085 @default.
- W4378647911 hasAuthorship W4378647911A5075724211 @default.
- W4378647911 hasBestOaLocation W43786479111 @default.
- W4378647911 hasConcept C105795698 @default.
- W4378647911 hasConcept C119043178 @default.
- W4378647911 hasConcept C119857082 @default.
- W4378647911 hasConcept C124101348 @default.
- W4378647911 hasConcept C154945302 @default.
- W4378647911 hasConcept C159620131 @default.
- W4378647911 hasConcept C169258074 @default.
- W4378647911 hasConcept C185592680 @default.
- W4378647911 hasConcept C198531522 @default.
- W4378647911 hasConcept C205649164 @default.
- W4378647911 hasConcept C27181475 @default.
- W4378647911 hasConcept C33923547 @default.
- W4378647911 hasConcept C41008148 @default.
- W4378647911 hasConcept C43617362 @default.
- W4378647911 hasConcept C58640448 @default.
- W4378647911 hasConcept C73555534 @default.
- W4378647911 hasConcept C92835128 @default.
- W4378647911 hasConcept C9770341 @default.
- W4378647911 hasConceptScore W4378647911C105795698 @default.
- W4378647911 hasConceptScore W4378647911C119043178 @default.
- W4378647911 hasConceptScore W4378647911C119857082 @default.
- W4378647911 hasConceptScore W4378647911C124101348 @default.
- W4378647911 hasConceptScore W4378647911C154945302 @default.
- W4378647911 hasConceptScore W4378647911C159620131 @default.
- W4378647911 hasConceptScore W4378647911C169258074 @default.
- W4378647911 hasConceptScore W4378647911C185592680 @default.
- W4378647911 hasConceptScore W4378647911C198531522 @default.
- W4378647911 hasConceptScore W4378647911C205649164 @default.
- W4378647911 hasConceptScore W4378647911C27181475 @default.
- W4378647911 hasConceptScore W4378647911C33923547 @default.
- W4378647911 hasConceptScore W4378647911C41008148 @default.
- W4378647911 hasConceptScore W4378647911C43617362 @default.
- W4378647911 hasConceptScore W4378647911C58640448 @default.
- W4378647911 hasConceptScore W4378647911C73555534 @default.
- W4378647911 hasConceptScore W4378647911C92835128 @default.
- W4378647911 hasConceptScore W4378647911C9770341 @default.
- W4378647911 hasFunder F4320322725 @default.
- W4378647911 hasLocation W43786479111 @default.
- W4378647911 hasOpenAccess W4378647911 @default.
- W4378647911 hasPrimaryLocation W43786479111 @default.
- W4378647911 hasRelatedWork W2020701082 @default.