Machine Learning and Instrument Autonomy Group

Science Serving Data Science

MLIA is leading JPL’s efforts to support scientists in Astronomy, Planetary Science, and Earth Science to gain insight and process-level understanding from large and complex data. Our approach on Data Science for Science Understanding (SUDS) is changing how research is performed across JPL and beyond.

Overview of data science for physical scientists by Dr. Lukas Mandrake at the JPL SUDS Seminar Series

Areas of Expertise

Science Insight from ML Interpretable Modeling
Large & complex data discovery Publication-supporting product quality & delivery
Catalog purification & exploration Explainable, justifiable model selection
Physical & ML model fusion Interpretable model behavior
Identification of key drivers Rigorous performance statistics
Bias discovery & correction Physical model acceleration
Uncertainty Awareness Global scalability / applicability

Current Projects

  • Bias
  • Scientific

Past Projects

  • Adaptive

  • Agricultural

  • MISR

  • Mars