Multi-ouput Gaussian Process Coregion models with Hadamard product #454
Conversation
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
danhphan
force-pushed
the
mogp-lcm-hadamard
branch
from
4fc71ad to
6decf20
Compare
|
View / edit / reply to this conversation on ReviewNB fonnesbeck commented on 2022-11-01T13:32:46Z "how to implement the Intrinsic..."
"Hadamard product between the Coregion kernel..."
"information about ICM..."
"please check out the talk..."
"by Mauricio Alvarez..."
"in different games..." |
|
View / edit / reply to this conversation on ReviewNB fonnesbeck commented on 2022-11-01T13:32:47Z If you add a |
|
View / edit / reply to this conversation on ReviewNB fonnesbeck commented on 2022-11-01T13:32:48Z Might add a line describing the advantage of LCM (or why you might want to use it over ICM) |
|
View / edit / reply to this conversation on ReviewNB fonnesbeck commented on 2022-11-01T13:32:49Z Perhaps bump tuning to 2000 to try and get rid of the divergences. |
|
Hi yes, thanks Chris, I will address these comments this weekend! |
danhphan
force-pushed
the
mogp-lcm-hadamard
branch
from
6decf20 to
4f50fda
Compare
|
Thank you Chris, I have updated the notebook according to your suggestions. Let me know if it needs any further changes. Also will need one more reviewer for this notebook. Thanks! |
|
View / edit / reply to this conversation on ReviewNB bwengals commented on 2022-11-07T23:19:41Z Though I don't mind being associated with this at all, I didn't write any of it! I think you deserve full credit here Danh, and Chris too via the data. danhphan commented on 2022-11-12T07:34:30Z Hi Bill, I would like to keep the authors like that. Since several parts of this notebook are from your previous codes of ICM on PyMC3. Besides, it makes sense to include the mentors, as they also involve a lot in the ideas and suggestions of the notebook's content. Thank you. |
|
View / edit / reply to this conversation on ReviewNB bwengals commented on 2022-11-07T23:19:42Z Same watermark as gp-kron,
%watermark -n -u -v -iv -w -p aesara,aeppl,xarray |
|
Hi Bill, I would like to keep the authors like that. Since several parts of this notebook are from your previous codes of ICM on PyMC3. Besides, it makes sense to include the mentors, as they also involve a lot in the ideas and suggestions of the notebook's content. Thank you. View entire conversation on ReviewNB |
danhphan
force-pushed
the
mogp-lcm-hadamard
branch
from
4f50fda to
a6a8601
Compare
|
Hi, thank you Bill, I have updated it :) |
danhphan
force-pushed
the
mogp-lcm-hadamard
branch
from
a6a8601 to
7090635
Compare
|
Thanks @OriolAbril. The notebook has been updated with reference and watermark :) |
|
Hi @OriolAbril @fonnesbeck If no further changes, this PR should be good to be merged. Thank you! |
|
Hi, thank you @fonnesbeck ! |
This PR adds an example of how to implement Intrinsic Coregionalization Model (ICM) and Linear Coregionalization Model (LCM) using a Hamadard product between Coregion kernel and input kernels.
We model the average spin rates of several pitchers on different game dates from a baseball dataset.
Cc @bwengals @fonnesbeck
Let me know if it needs to update anythings.
Thank you.