Intelligent Automated Data Analysis using Advanced Machine Learning Algorithms

 

Research Staff: Maura Rabbette and Peter Pilewskie

The goal of this project is to develop a software framework incorporating advanced machine learning algorithms with visualization routines for the analysis of time-series data such as Solar Spectral Irradiance Measurements as recorded by the NASA Ames Solar Spectral Flux Radiometer (SSFR). The SSFR was developed for the Radiation Science Program of the Earth Science Enterprise and has been integrated on numerous aircraft to measure both upwelling and downwelling spectral solar irradiance. The SSFR has several hundred channels over the spectral range 300 - 2200 nm.

The objective is to develop new automated data reduction and analysis methods using advanced machine learning techniques. This will include further development of retrieval algorithms and on-board data processing capabilities that will be necessary for future "smart" sensors. These algorithms will be developed for both stationary and non-stationary time-series and will include hidden Markov models, neural networks, Principal Component Analysis (PCA), Independent Component Analysis (ICA) and Bayesian techniques.

These developments will be essential for future remote sensing experiments and onboard reduction of large volumes of satellite hyperspectra, for performing instrument diagnostics, instrument validation, for relating measured spectral variability to physical causes, and ultimately constraining climate models.

 

Collaborators: Kevin Wheeler and Dogan Timucin, Code IC

Point of Contact: Maura Rabbette, (650/604-0128, mrabbette@mail.arc.nasa.gov