-
What should be our long term pattern for hyper parameter selection?
examples: lag selection in tsa models, penalty in penalized models.
tradeoff: clean, more "fool-proof" design versus computati…
-
This issue is mainly to link to several comments and issues where this shows up.
This shows up in penalized estimation.
But it is related according to Wood on splines in R book also to mixed effec…
-
default methods to find linear combos of predictors in classification and regression trees
-
currently we treat zero residuals as positive
after fixing #2377 zeros are moved to the positive part, then they are multiplied by (1-q) in the `where`, which already had a weak inequality.
```
mask…
-
@tlienart
1. **All Subset (Best K Subset)**:
I've got parsimonious code for All Subset Regression which might have a nice home in this package that I'd like to share.
While it's much faster than …
-
### Describe the workflow you want to enable
I'd like to have a feature importance method native to linear models (without L1 penalty) that is calculated on the training set:
```python
clf = Logi…
-
question on mailing list
general penalized discussions in #2282, #2293
This is a specific issue for implementing Firth's bias correcting penalization, which is different from the L1, L2, CAD penal…
-
Just wondering if it would be possible to support an extra argument weights with observation weights? Right now I can fit an `L0Learn `model with observation weights by multiplying both my covariate m…
-
Dear Prof @bcjaeger thank you so much for such a great package.
First, Here is my two cents on improving the speed of `method='net'`.
In [penalized_cph.R](https://github.com/ropensci/aorsf/blob…
-
A nice interface will only be useful if people actually implement it. I think this will by far be the major hurdle to overcome.
Here's some infrastructure that I believe in necessary for interface …