-
Looking at the [annotation model](https://aclanthology.org/Q14-1025.pdf), first thing that needs to be added are multidimensional parameters. I've thought about how we _might_ try to make it work with…
-
See [here](https://sethrf.com/files/fast-hierarchical-GPs.pdf). The idea is to exploit separable structure rather than dealing with one big covariance matrix. Would this be something worth implementin…
-
Hi,
Thanks for this package. i really like its concept but I'm still not clear on some things.
I am working on the following code for a multi-response LMM:
```
using Distributions
using…
-
Song Y, Song J, Ermon S. [Accelerating natural gradient with higher-order invariance](http://proceedings.mlr.press/v80/song18a/song18a.pdf).
-
The implementation of `UniGrid` is wrong at the moment it should return a smart matrix object but just returns a vector of range right now
-
Consider the following code sample:
function type_testB(B::StridedArray)
return typeof(similar(B, size(B))) …
-
Update the following URL to point to the GitHub repository of
the package you wish to submit to _Bioconductor_
- Repository: https://github.com/tidymass/masstools
Confirm the following by editi…
-
I am trying to implement a version of the log-determinant approximation given in [this](https://arxiv.org/pdf/1310.5288.pdf) paper.
I am implementing a model very similar to the one in this paper. …
-
Hi,
I've been using the neural-tangents library a lot over the past few months, it's been extremely helpful.
I just a had a question about calculating the marginal log-likelihood for NNGPs, wh…
-
Right now we save dense covariance matrices, but when the dimension is high enough this is a bad idea. In fact, to solve a PDE I locally modified the `ODEFilter.solve` function to only save the Kronec…