Mitigating Imbalance of Land Cover Change Data for Deep Learning Models with Temporal and Spatiotemporal Sample Weighting Schemes
An open problem impeding the use of deep learning (DL) models for forecasting land cover (LC) changes is their bias toward persistent cells. By providing sample weights for model training, LC changes can be allocated greater influence in adjustments to model internal parameters. The main goal of thi...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-11-01
|
| Series: | ISPRS International Journal of Geo-Information |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2220-9964/11/12/587 |
| _version_ | 1827638453352792064 |
|---|---|
| author | Alysha van Duynhoven Suzana Dragićević |
| author_facet | Alysha van Duynhoven Suzana Dragićević |
| author_sort | Alysha van Duynhoven |
| collection | DOAJ |
| description | An open problem impeding the use of deep learning (DL) models for forecasting land cover (LC) changes is their bias toward persistent cells. By providing sample weights for model training, LC changes can be allocated greater influence in adjustments to model internal parameters. The main goal of this research study was to implement and evaluate temporal and spatiotemporal sample weighting schemes that manage the influence of persistent and formerly changed areas. The proposed sample weighting schemes allocate higher weights to more recently changed areas based on the inverse temporal and spatiotemporal distance from previous changes occurring at a location or within the location’s neighborhood. Four spatiotemporal DL models (CNN-LSTM, CNN-GRU, CNN-TCN, and ConvLSTM) were used to compare the sample weighting schemes to forecast the LC changes of the Columbia-Shuswap Regional District in British Columbia, Canada, using data obtained from the MODIS annual LC dataset and other auxiliary spatial variables. The results indicate that the presented weighting schemes facilitated improvement over no sample weighting and the common inverse frequency weighting scheme for multi-year LC change forecasts, lowering errors due to quantity while reducing overall allocation error severity. This research study contributes to strategies for addressing the characteristic imbalances of multitemporal LC change datasets for DL modeling endeavors. |
| first_indexed | 2024-03-09T16:20:32Z |
| format | Article |
| id | doaj.art-42cd308488f94332864ab4a27475e998 |
| institution | Directory Open Access Journal |
| issn | 2220-9964 |
| language | English |
| last_indexed | 2024-03-09T16:20:32Z |
| publishDate | 2022-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | ISPRS International Journal of Geo-Information |
| spelling | doaj.art-42cd308488f94332864ab4a27475e9982023-11-24T15:20:46ZengMDPI AGISPRS International Journal of Geo-Information2220-99642022-11-01111258710.3390/ijgi11120587Mitigating Imbalance of Land Cover Change Data for Deep Learning Models with Temporal and Spatiotemporal Sample Weighting SchemesAlysha van Duynhoven0Suzana Dragićević1Spatial Analysis and Modeling Laboratory, Department of Geography, Simon Fraser University, 8888 University Drive, Burnaby, BC V5A1S6, CanadaSpatial Analysis and Modeling Laboratory, Department of Geography, Simon Fraser University, 8888 University Drive, Burnaby, BC V5A1S6, CanadaAn open problem impeding the use of deep learning (DL) models for forecasting land cover (LC) changes is their bias toward persistent cells. By providing sample weights for model training, LC changes can be allocated greater influence in adjustments to model internal parameters. The main goal of this research study was to implement and evaluate temporal and spatiotemporal sample weighting schemes that manage the influence of persistent and formerly changed areas. The proposed sample weighting schemes allocate higher weights to more recently changed areas based on the inverse temporal and spatiotemporal distance from previous changes occurring at a location or within the location’s neighborhood. Four spatiotemporal DL models (CNN-LSTM, CNN-GRU, CNN-TCN, and ConvLSTM) were used to compare the sample weighting schemes to forecast the LC changes of the Columbia-Shuswap Regional District in British Columbia, Canada, using data obtained from the MODIS annual LC dataset and other auxiliary spatial variables. The results indicate that the presented weighting schemes facilitated improvement over no sample weighting and the common inverse frequency weighting scheme for multi-year LC change forecasts, lowering errors due to quantity while reducing overall allocation error severity. This research study contributes to strategies for addressing the characteristic imbalances of multitemporal LC change datasets for DL modeling endeavors.https://www.mdpi.com/2220-9964/11/12/587land cover changespatiotemporal deep learninggeospatial data imbalancesample weightsinverse temporal distance weightingspatiotemporal distance weighting |
| spellingShingle | Alysha van Duynhoven Suzana Dragićević Mitigating Imbalance of Land Cover Change Data for Deep Learning Models with Temporal and Spatiotemporal Sample Weighting Schemes ISPRS International Journal of Geo-Information land cover change spatiotemporal deep learning geospatial data imbalance sample weights inverse temporal distance weighting spatiotemporal distance weighting |
| title | Mitigating Imbalance of Land Cover Change Data for Deep Learning Models with Temporal and Spatiotemporal Sample Weighting Schemes |
| title_full | Mitigating Imbalance of Land Cover Change Data for Deep Learning Models with Temporal and Spatiotemporal Sample Weighting Schemes |
| title_fullStr | Mitigating Imbalance of Land Cover Change Data for Deep Learning Models with Temporal and Spatiotemporal Sample Weighting Schemes |
| title_full_unstemmed | Mitigating Imbalance of Land Cover Change Data for Deep Learning Models with Temporal and Spatiotemporal Sample Weighting Schemes |
| title_short | Mitigating Imbalance of Land Cover Change Data for Deep Learning Models with Temporal and Spatiotemporal Sample Weighting Schemes |
| title_sort | mitigating imbalance of land cover change data for deep learning models with temporal and spatiotemporal sample weighting schemes |
| topic | land cover change spatiotemporal deep learning geospatial data imbalance sample weights inverse temporal distance weighting spatiotemporal distance weighting |
| url | https://www.mdpi.com/2220-9964/11/12/587 |
| work_keys_str_mv | AT alyshavanduynhoven mitigatingimbalanceoflandcoverchangedatafordeeplearningmodelswithtemporalandspatiotemporalsampleweightingschemes AT suzanadragicevic mitigatingimbalanceoflandcoverchangedatafordeeplearningmodelswithtemporalandspatiotemporalsampleweightingschemes |