Rockerstop
Thesis project
The project proposes a new method to make climate models more adaptive and accurate using reinforcement learning and federated learning. Traditional climate models rely on fixed, manually tuned parameters to represent physical processes such as radiation. These parameters are calibrated offline and remain constant, even though climate conditions change over time. As a result, models may become inaccurate under new or evolving climate states.
To address this limitation, the authors introduce a f
More Details:
Lucknow
People contacted: Signup to view
Don't Miss any jobs & Projects
Join Rockerstop Today to get "email alert" It`s Free !