New answers tagged scikit-learn
0
votes
Classification Threshold Tuning with GridSearchCV
You can use TunedThresholdClassifierCV to find the best hyperparameters with cross-validation.
And you can keep other parametrs fixed or not.
- 185
0
votes
As an intermediate R programmer looking to dive into machine learning, should I choose Python or stick with R?
R statistics package is a large ecosystem of tools, purpose built and actively maintained for statistics by statisticians, for accomplishing a variety of statistics, machine learning and many other ...
- 2,565
1
vote
How are the successive sets of training samples that are allocated for each iteration of HalvingGridSearchCV determined?
My understanding from looking at the source code is that the samples are selected at random. The code is here, and this class is used later in the successive halving here.
- 291
0
votes
Multivariate linear regression via scikit and statsmodels
This seems sensible. If you look at some of the formulas for the fitted regression coefficients, $y$ and $\beta$ turn from a vectors into matrices. The computations are such that the columns of $B$ (...
- 156
0
votes
Splitter in decision trees in sklearn implementation
What is the benefit of splitter = 'random' over splitter = 'best' besides training speed? And how much of a boost is that choice ...
0
votes
My results from linear regression differs from my collegues despite having same data. Is this to be expected?
OLS linear regression is deterministic (insofar as computers are deterministic). There is a straightforward calculation to obtain the parameter estimates. In that regard, yes, you should obtain the ...
- 4,044
0
votes
My results from linear regression differs from my collegues despite having same data. Is this to be expected?
TBH, this is going to be tough, as we do not know what preprocessing step(s) was done. And if this was not enough, we don't even know for sure if the model was really a standard linear regressor. Let'...
- 1,117
Top 50 recent answers are included