Impact Assessment of Missing Data in Model Predictions for Earth Observation Applications

Earth observation (EO) applications involving complex and heterogeneous data sources are commonly approached with machine learning models. However, there is a common assumption that data sources will be persistently available. Different situations could affect the availability of EO sources, like no...

Full description

Saved in:
Bibliographic Details
Published inIEEE International Geoscience and Remote Sensing Symposium proceedings pp. 967 - 971
Main Authors Mena, Francisco, Arenas, Diego, Charfuelan, Marcela, Nuske, Marlon, Dengel, Andreas
Format Conference Proceeding
LanguageEnglish
Published IEEE 07.07.2024
Subjects
Data models
Earth
Missing Information
Multi-view Learning
Noise
Predictive models
Satellites
Soft sensors
Time series analysis
Time Series Data
Vegetation Science
Online AccessGet full text

Cover

More Information
Summary:Earth observation (EO) applications involving complex and heterogeneous data sources are commonly approached with machine learning models. However, there is a common assumption that data sources will be persistently available. Different situations could affect the availability of EO sources, like noise, clouds, or satellite mission failures. In this work, we assess the impact of missing temporal and static EO sources in trained models across four datasets with classification and regression tasks. We compare the predictive quality of different methods and find that some are naturally more robust to missing data. The Ensemble strategy, in particular, achieves a prediction robustness up to 100%. We evidence that missing scenarios are significantly more challenging in regression than classification tasks. Finally, we find that the optical view is the most critical view when it is missing individually.
ISSN:2153-7003
DOI:10.1109/IGARSS53475.2024.10640375