Thank you.
I was however looking for whether the developers intend to implement sparse
matrix support for GBRT
On Sun, Aug 4, 2013 at 10:21 PM, Maheshakya Wijewardena <
pmaheshak...@gmail.com> wrote:
> As far I know. these are the only models that support sparse matrices
>
> linear_model.Logistic
2013/8/4 Maheshakya Wijewardena :
> On Sun, Aug 4, 2013 at 7:34 PM, hrishi wrote:
>>
>> Are there any plans to implement sparse arrays for gradient boosted
>> regression trees in scikit?
Plans, yes, but no time. The recent refactoring of the trees module
ought to have made this easier; a patch is
As far I know. these are the only models that support sparse matrices
linear_model.LogisticRegression()
svm.SVR()
svm.NuSVR()
linear_model.LinearRegression()
neighbors.KNeighborsRegressor()
naive_bayes.MultinomialNB()
naive_bayes.BernoulliNB()
linear_model.PassiveAggressiveRegressor()
linear_model
My data consists solely of categorical variables and I have used one hot
encoding.
I can't convert the sparse matrix to dense since I have too many categories.
Are there any plans to implement sparse arrays for gradient boosted
regression trees in scikit?
--
Hrishikesh V. Ganu
Mobile: 9740639
Hi everybody,
I'm currently working on a Pull Request for Gradient Boosted
Regression Trees [1] (aka Gradient Boosting, MART, TreeNet) and I'm
looking for collaborators.
GBRTs have been advertised as one of the best off-the-shelf
data-mining procedures; they share many properties with random fore
