Geo-information Science and Remote Sensing, Wageningen University, The Netherlands
Deadline: Feb 15, 2019
A two-year research position on near real-time forest monitoring using Sentinel satellite data and machine learning.
Further innovations are now needed to utilize the high spatial and temporal information content of the new satellite missions for improving the detection and characterization of changes, and for scaling the developed approaches. Solutions based on time-dependent deep learning models (e.g. GRU, LSTM) will be explored, with a particular focus on the inclusion of spatial awareness and robustness to missing data.
- motivated researcher with a data science, machine learning and/or geoinformatics degree.
- Experience with machine learning methods
- handling large satellite time series data (e.g. Google Earth Engine)
- programming skills (e.g. R, Python, java script)
- Knowledge about radar remote sensing
- Understanding of tropical forest dynamics is an advantage.
- Strong English skills are fundamental.