Revisiting the Modifiable Areal Unit Problem in Deep Traffic Prediction with Visual Analytics

Deep learning methods are being increasingly used for urban traffic prediction where spatiotemporal traffic data is aggregated into sequentially organized matrices that are then fed into convolution-based residual neural networks. However, the widely known modifiable areal unit problem within such a...

Description complète

Détails bibliographiques
Publié dans:IEEE transactions on visualization and computer graphics. - 1998. - 27(2021), 2 vom: 28. Feb., Seite 839-848
Auteur principal: Zeng, Wei (Auteur)
Autres auteurs: Lin, Chengqiao, Lin, Juncong, Jiang, Jincheng, Xia, Jiazhi, Turkay, Cagatay, Chen, Wei
Format: Article en ligne
Langue:English
Publié: 2021
Accès à la collection:IEEE transactions on visualization and computer graphics
Sujets:Journal Article
LEADER 01000caa a22002652c 4500
001 NLM316441759
003 DE-627
005 20250228054314.0
007 cr uuu---uuuuu
008 231225s2021 xx |||||o 00| ||eng c
024 7 |a 10.1109/TVCG.2020.3030410  |2 doi 
028 5 2 |a pubmed25n1054.xml 
035 |a (DE-627)NLM316441759 
035 |a (NLM)33074818 
040 |a DE-627  |b ger  |c DE-627  |e rakwb 
041 |a eng 
100 1 |a Zeng, Wei  |e verfasserin  |4 aut 
245 1 0 |a Revisiting the Modifiable Areal Unit Problem in Deep Traffic Prediction with Visual Analytics 
264 1 |c 2021 
336 |a Text  |b txt  |2 rdacontent 
337 |a ƒaComputermedien  |b c  |2 rdamedia 
338 |a ƒa Online-Ressource  |b cr  |2 rdacarrier 
500 |a Date Revised 02.02.2021 
500 |a published: Print-Electronic 
500 |a Citation Status PubMed-not-MEDLINE 
520 |a Deep learning methods are being increasingly used for urban traffic prediction where spatiotemporal traffic data is aggregated into sequentially organized matrices that are then fed into convolution-based residual neural networks. However, the widely known modifiable areal unit problem within such aggregation processes can lead to perturbations in the network inputs. This issue can significantly destabilize the feature embeddings and the predictions - rendering deep networks much less useful for the experts. This paper approaches this challenge by leveraging unit visualization techniques that enable the investigation of many-to-many relationships between dynamically varied multi-scalar aggregations of urban traffic data and neural network predictions. Through regular exchanges with a domain expert, we design and develop a visual analytics solution that integrates 1) a Bivariate Map equipped with an advanced bivariate colormap to simultaneously depict input traffic and prediction errors across space, 2) a Moran's I Scatterplot that provides local indicators of spatial association analysis, and 3) a Multi-scale Attribution View that arranges non-linear dot plots in a tree layout to promote model analysis and comparison across scales. We evaluate our approach through a series of case studies involving a real-world dataset of Shenzhen taxi trips, and through interviews with domain experts. We observe that geographical scale variations have important impact on prediction performances, and interactive visual exploration of dynamically varying inputs and outputs benefit experts in the development of deep traffic prediction models 
650 4 |a Journal Article 
700 1 |a Lin, Chengqiao  |e verfasserin  |4 aut 
700 1 |a Lin, Juncong  |e verfasserin  |4 aut 
700 1 |a Jiang, Jincheng  |e verfasserin  |4 aut 
700 1 |a Xia, Jiazhi  |e verfasserin  |4 aut 
700 1 |a Turkay, Cagatay  |e verfasserin  |4 aut 
700 1 |a Chen, Wei  |e verfasserin  |4 aut 
773 0 8 |i Enthalten in  |t IEEE transactions on visualization and computer graphics  |d 1998  |g 27(2021), 2 vom: 28. Feb., Seite 839-848  |w (DE-627)NLM098269445  |x 1941-0506  |7 nnas 
773 1 8 |g volume:27  |g year:2021  |g number:2  |g day:28  |g month:02  |g pages:839-848 
856 4 0 |u http://dx.doi.org/10.1109/TVCG.2020.3030410  |3 Volltext 
912 |a GBV_USEFLAG_A 
912 |a SYSFLAG_A 
912 |a GBV_NLM 
912 |a GBV_ILN_350 
951 |a AR 
952 |d 27  |j 2021  |e 2  |b 28  |c 02  |h 839-848