esem.utils.leave_one_out¶
- esem.utils.leave_one_out(Xdata, Ydata, model='RandomForest', **model_kwargs)¶
Function to perform LeaveOneOut cross-validation with different models.
- Parameters
Xdata (array-like of
shape (n_samples
,n_features)
) – Parameter valuesYdata (array-like of
shape (n_samples,)
) – Target values.model (
{'RandomForest', 'GaussianProcess', 'NeuralNet'}
, default'RandomForest'
)model_kwargs (
dict
) – More arguments to pass to the model.
- Returns
output (
list
ofn_samples (truth
,prediction
,variance) tuples
) – which can then be passed to esem.utils.validation_plot()