Приказ основних података о документу

dc.creatorNikezić, Dušan P.
dc.creatorRadivojević, Dušan S.
dc.creatorLazović, Ivan
dc.creatorMirkov, Nikola S.
dc.creatorMarković, Zoran J.
dc.date.accessioned2024-03-15T07:17:05Z
dc.date.available2024-03-15T07:17:05Z
dc.date.issued2024
dc.identifier.issn2227-7390
dc.identifier.urihttps://vinar.vin.bg.ac.rs/handle/123456789/12988
dc.description.abstractIn order to better predict the high aerosol concentrations associated with air pollution and climate change, a machine learning model was developed using transfer learning and the segmentation process of global satellite images. The main concept of transfer learning lies on convolutional neural networks and works by initializing the already trained model weights to better adapt the weights when the network is trained on a different dataset. The transfer learning technique was tested with the ResNet3D-101 model pre-trained from a 2D ImageNet dataset. This model has performed well for contrail detection to assess climate impact. Aerosol distributions can be monitored via satellite remote sensing. Satellites can monitor some aerosol optical properties like aerosol optical thickness. Aerosol optical thickness snapshots were the input dataset for the model and were obtained from NASA’s Terra-Modis satellite; the output images were segmented by comparing the pixel values with a threshold value of 0.8 for aerosol optical thickness. Hyperparameter optimization finds a tuple of hyperparameters that yields an optimal model that minimizes a predefined loss function on given independent data. The model structure was adjusted in order to improve the performance of the model by applying methods and hyperparameter optimization techniques such as grid search, batch size, threshold, and input length. According to the criteria defined by the authors, the distance domain criterion and time domain criterion, the developed model is capable of generating adequate data and finding patterns in the time domain. As observed from the comparison of relative coefficients for the criteria metrics proposed by the authors, ddc and dtc, the deep learning model based on ConvLSTM layers developed in our previous studies has better performance than the model developed in this study with transfer learning.en
dc.relationinfo:eu-repo/grantAgreement/MESTD/inst-2020/200017/RS//
dc.rightsopenAccess
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.sourceMathematicsen
dc.subjecttransfer learningen
dc.subjectResNet3D-101en
dc.subjectaerosol optical thicknessen
dc.subjectdistance and time domain criteriaen
dc.subjectearly warning systemen
dc.titleTransfer Learning with ResNet3D-101 for Global Prediction of High Aerosol Concentrationsen
dc.typearticleen
dc.rights.licenseBY
dc.citation.volume12
dc.citation.issue6
dc.citation.spage826
dc.identifier.doi10.3390/math12060826
dc.type.versionpublishedVersion
dc.identifier.fulltexthttp://vinar.vin.bg.ac.rs/bitstream/id/35913/mathematics-12-00826-v2.pdf


Документи

Thumbnail

Овај документ се појављује у следећим колекцијама

Приказ основних података о документу