PyMC3 contains a rich suite of building blocks for probabilistic modeling and inference. This proposal contains two parts: first is finishing support for single precision, and second involves finishing Gaussian Process functionality. I have already submitted a proposal about exploring alternative computation engines, and am open to working with the mentors on mixing and matching projects from each, for example, evaluating PyTorch in addition to the projects in this proposal.

Organization

Student

Bill Engels

Mentors

  • Chris Fonnesbeck
  • Austin Rochford
close

2017