parallel-SSVI (WIP)

The code is heavily borrowed from Newt.

PSSGP:

TODO

• Define Model
• Matern Kernel
• Kernel Utils
• Balancing
• Lyap eq solver
• Sequential Kalman Filter
• Parallel Kalman Filter
• JIT Kalman Filters
• Test Matern Kernel with GpyTorch
• Compare GPR with sequential implementation.

TODO:

• State Space GP Model
  • GH quadrature
  • Kalman Filter
  • RTS Smoother
• Kernels
  • Matern32
  • Matern52
• Associative scan
• Datasets

Getting Started

References

• https://github.com/EEA-sensors/parallel-gps
• https://github.com/AaltoML/Newt
• https://github.com/AaltoML/kalman-jax