Machine Learning Applications to Blending Challenges
Wednesday, 1:30 - 3:00pm Pacific, Coronado II.
- Overview of ongoing projects using machine learning to address blending challenges (detecting, identifying, deblending, or measuring blended objects) with a focus on a set of common organizing questions.
- Overview of tools for training & testing algorithms, including performance metrics; potential for ML applications in image generation.
- Perspective from LSST DM (Jim Bosch, science pipelines).
- Identification and discussion of promising approaches and common issues that could be explored or addressed more effectively across science collaborations.
All presented material is in this set of google slides and the attached pdf (linked below).
Google slides are now shared so that anyone with the link can comment. If you need edit access, please request it through the google slides or by contacting Pat Burchat [burchat AT stanford DOT edu].