Christos Chatzis

PARAFAC2 for missing data

1

Hi! I am currently working with PARAFAC2 and I noticed there is no support for missing data (PARAFAC has the `mask` argument). Are there any plans for missing data support...

Factor tools for PARAFAC2

2

Hi there, thank you making my life easier with tlviz. I was wondering, would there be interest in updates of functions in `factor_tools.py` to support `PARAFAC2Tensor`s?

Parafac2 als em

7

Enable missing data imputation via the `mask` argument as done in CP, as brought up in #543 . The changes are outlined as follows (all lines refer to `_parafac2.py`): 1....

Hi, as mentioned in #23, here is an example notebook with the replicability check approach for selecting the number of components. Kindly let me know what you think.

Replicability as a model evaluation technique

1

Hi! Would it be of interest to have replicability (i.e. are the components discovered replicable?) as another metric to evaluate cp models? The replicability check would boil down to the...

Hi! Me and @blzserdos have a notebook describing the [replicability procedure](https://github.com/tensorly/viz/issues/23) for choosing the number of components for PARAFAC2. We already made a [PR](https://github.com/tensorly/viz/pull/24) for this check for CP models...

Ridge regularization for PARAFAC2

2

Hi! I noticed there does not exist an `l2_reg` penalty argument to include ridge regularization on all factors for PARAFAC2 like in CP. If you also think this would be...