SciML / PDERoadmap

A repository for the discussion of PDE tooling for scientific machine learning (SciML) and physics-informed machine learning
https://tutorials.sciml.ai/
18 stars 6 forks source link

When does/doesn't the drift/variance of a diffusion enter the Q? #19

Closed jlperla closed 6 years ago

jlperla commented 6 years ago

After we undestand #18 we should make sure we understand why the drift and variance didn't enter the Q for the jump-diffusion example in the document (or, if there is a mistake, fix it!)

While it is clear that those should not enter the B matrix, I don't see why they disappear in the Gaussian Elimination? Does this still work if there is variation in the drifts?

jlperla commented 6 years ago

Discussed with @stevenzhangdx . Looks like it just cancels out perfectly with the diffusion process with reflecting barriers, and the cancellation will work for state varying as well. Not quite so clear on what would happen with a non-zero absorbing barrier, but we can wait on that.