SSAIHQ CMS Display Portlet
Accelerating Employee Growth with SSAI’s Deep Learning Academy
SSAI’s dedication to helping employees grow in high demand areas has led us to invest in an accelerator program designed to significantly improve skills within the machine learning field in a short amount of time.
Essentially, each group of learners in SSAI’s Deep Learning Academy (DLA) is given a set of projects that apply machine learning concepts to unsolved, real-world problems. In this way, learners are integrated into a realistic, applied science environment, and exit the program with a firm understanding of how their new knowledge in machine and deep learning can be applied to their own work—particularly with respect to issues and questions relevant to NASA and other partners.
To help motivate innovative solution development in a fun, collaborative manner, the learners are grouped into teams, and compete to solve the problem sets, while earning prizes along the way. Solutions are shared with SSAI’s customers, helping to advance knowledge and innovation in difficult problem areas.
One of our current learners, Ranjay Shrestha, plans to put his machine learning skills to use in support of NASA’s Black Marble project:
"I am planning to utilize Machine Learning approaches I learned through this Deep Learning Specialization course to downscale VIIRS derived Black Marble Nighttime Light product by identifying the distribution of the intensity of the light within urban land use pixels and improve current VIIRS cloud/cloud shadow detection algorithms. "
Another learner, Steve Cox, who currently supports the NASA Airborne Science program at NASA Langley Research Center, plans to apply machine learning techniques “to improve the accuracy of NASA/GEWEX Surface Radiation Budget products, especially in areas of missing satellite coverage.”
Moving forward, the DLA will include two tracks—one focused on data science and the other on Machine Learning Operations (MLOps). The data science track will build on the current program capabilities, allowing learners to improve their understanding of the machine learning algorithmic design and neural network creation processes. The MLOps track will offer learners the opportunity to gain the required skills for implementing models into production, as well as an understanding of what engineering considerations go into designing a model within a hardware-constrained environment (e.g. edge computing, embedded ML).
Once the curriculum is fully developed, learners from these different tracks will pair up to comprise project teams. This learning approach enables participants to practice problem solving in a simulated, real-world environment, where science, engineering, and cross-discipline collaboration are key to mission success.