3.2.4.3.5. sklearn.ensemble
.GradientBoostingClassifier¶

class
sklearn.ensemble.
GradientBoostingClassifier
(loss=’deviance’, learning_rate=0.1, n_estimators=100, subsample=1.0, criterion=’friedman_mse’, min_samples_split=2, min_samples_leaf=1, min_weight_fraction_leaf=0.0, max_depth=3, min_impurity_decrease=0.0, min_impurity_split=None, init=None, random_state=None, max_features=None, verbose=0, max_leaf_nodes=None, warm_start=False, presort=’auto’, validation_fraction=0.1, n_iter_no_change=None, tol=0.0001)[source]¶ Gradient Boosting for classification.
GB builds an additive model in a forward stagewise fashion; it allows for the optimization of arbitrary differentiable loss functions. In each stage
n_classes_
regression trees are fit on the negative gradient of the binomial or multinomial deviance loss function. Binary classification is a special case where only a single regression tree is induced.Read more in the User Guide.
Parameters:  loss : {‘deviance’, ‘exponential’}, optional (default=’deviance’)
loss function to be optimized. ‘deviance’ refers to deviance (= logistic regression) for classification with probabilistic outputs. For loss ‘exponential’ gradient boosting recovers the AdaBoost algorithm.
 learning_rate : float, optional (default=0.1)
learning rate shrinks the contribution of each tree by
learning_rate
. There is a tradeoff between learning_rate and n_estimators. n_estimators : int (default=100)
The number of boosting stages to perform. Gradient boosting is fairly robust to overfitting so a large number usually results in better performance.
 subsample : float, optional (default=1.0)
The fraction of samples to be used for fitting the individual base learners. If smaller than 1.0 this results in Stochastic Gradient Boosting.
subsample
interacts with the parametern_estimators
. Choosingsubsample < 1.0
leads to a reduction of variance and an increase in bias. criterion : string, optional (default=”friedman_mse”)
The function to measure the quality of a split. Supported criteria are “friedman_mse” for the mean squared error with improvement score by Friedman, “mse” for mean squared error, and “mae” for the mean absolute error. The default value of “friedman_mse” is generally the best as it can provide a better approximation in some cases.
New in version 0.18.
 min_samples_split : int, float, optional (default=2)
The minimum number of samples required to split an internal node:
 If int, then consider
min_samples_split
as the minimum number.  If float, then
min_samples_split
is a fraction andceil(min_samples_split * n_samples)
are the minimum number of samples for each split.
Changed in version 0.18: Added float values for fractions.
 If int, then consider
 min_samples_leaf : int, float, optional (default=1)
The minimum number of samples required to be at a leaf node. A split point at any depth will only be considered if it leaves at least
min_samples_leaf
training samples in each of the left and right branches. This may have the effect of smoothing the model, especially in regression. If int, then consider
min_samples_leaf
as the minimum number.  If float, then
min_samples_leaf
is a fraction andceil(min_samples_leaf * n_samples)
are the minimum number of samples for each node.
Changed in version 0.18: Added float values for fractions.
 If int, then consider
 min_weight_fraction_leaf : float, optional (default=0.)
The minimum weighted fraction of the sum total of weights (of all the input samples) required to be at a leaf node. Samples have equal weight when sample_weight is not provided.
 max_depth : integer, optional (default=3)
maximum depth of the individual regression estimators. The maximum depth limits the number of nodes in the tree. Tune this parameter for best performance; the best value depends on the interaction of the input variables.
 min_impurity_decrease : float, optional (default=0.)
A node will be split if this split induces a decrease of the impurity greater than or equal to this value.
The weighted impurity decrease equation is the following:
N_t / N * (impurity  N_t_R / N_t * right_impurity  N_t_L / N_t * left_impurity)
where
N
is the total number of samples,N_t
is the number of samples at the current node,N_t_L
is the number of samples in the left child, andN_t_R
is the number of samples in the right child.N
,N_t
,N_t_R
andN_t_L
all refer to the weighted sum, ifsample_weight
is passed.New in version 0.19.
 min_impurity_split : float, (default=1e7)
Threshold for early stopping in tree growth. A node will split if its impurity is above the threshold, otherwise it is a leaf.
Deprecated since version 0.19:
min_impurity_split
has been deprecated in favor ofmin_impurity_decrease
in 0.19. The default value ofmin_impurity_split
will change from 1e7 to 0 in 0.23 and it will be removed in 0.25. Usemin_impurity_decrease
instead. init : estimator or ‘zero’, optional (default=None)
An estimator object that is used to compute the initial predictions.
init
has to providefit
andpredict_proba
. If ‘zero’, the initial raw predictions are set to zero. By default, aDummyEstimator
predicting the classes priors is used. random_state : int, RandomState instance or None, optional (default=None)
If int, random_state is the seed used by the random number generator; If RandomState instance, random_state is the random number generator; If None, the random number generator is the RandomState instance used by
np.random
. max_features : int, float, string or None, optional (default=None)
The number of features to consider when looking for the best split:
 If int, then consider
max_features
features at each split.  If float, then
max_features
is a fraction andint(max_features * n_features)
features are considered at each split.  If “auto”, then
max_features=sqrt(n_features)
.  If “sqrt”, then
max_features=sqrt(n_features)
.  If “log2”, then
max_features=log2(n_features)
.  If None, then
max_features=n_features
.
Choosing
max_features < n_features
leads to a reduction of variance and an increase in bias.Note: the search for a split does not stop until at least one valid partition of the node samples is found, even if it requires to effectively inspect more than
max_features
features. If int, then consider
 verbose : int, default: 0
Enable verbose output. If 1 then it prints progress and performance once in a while (the more trees the lower the frequency). If greater than 1 then it prints progress and performance for every tree.
 max_leaf_nodes : int or None, optional (default=None)
Grow trees with
max_leaf_nodes
in bestfirst fashion. Best nodes are defined as relative reduction in impurity. If None then unlimited number of leaf nodes. warm_start : bool, default: False
When set to
True
, reuse the solution of the previous call to fit and add more estimators to the ensemble, otherwise, just erase the previous solution. See the Glossary. presort : bool or ‘auto’, optional (default=’auto’)
Whether to presort the data to speed up the finding of best splits in fitting. Auto mode by default will use presorting on dense data and default to normal sorting on sparse data. Setting presort to true on sparse data will raise an error.
New in version 0.17: presort parameter.
 validation_fraction : float, optional, default 0.1
The proportion of training data to set aside as validation set for early stopping. Must be between 0 and 1. Only used if
n_iter_no_change
is set to an integer.New in version 0.20.
 n_iter_no_change : int, default None
n_iter_no_change
is used to decide if early stopping will be used to terminate training when validation score is not improving. By default it is set to None to disable early stopping. If set to a number, it will set asidevalidation_fraction
size of the training data as validation and terminate training when validation score is not improving in all of the previousn_iter_no_change
numbers of iterations. The split is stratified.New in version 0.20.
 tol : float, optional, default 1e4
Tolerance for the early stopping. When the loss is not improving by at least tol for
n_iter_no_change
iterations (if set to a number), the training stops.New in version 0.20.
Attributes:  n_estimators_ : int
The number of estimators as selected by early stopping (if
n_iter_no_change
is specified). Otherwise it is set ton_estimators
.New in version 0.20.
feature_importances_
: array, shape (n_features,)Return the feature importances (the higher, the more important the feature).
 oob_improvement_ : array, shape (n_estimators,)
The improvement in loss (= deviance) on the outofbag samples relative to the previous iteration.
oob_improvement_[0]
is the improvement in loss of the first stage over theinit
estimator. train_score_ : array, shape (n_estimators,)
The ith score
train_score_[i]
is the deviance (= loss) of the model at iterationi
on the inbag sample. Ifsubsample == 1
this is the deviance on the training data. loss_ : LossFunction
The concrete
LossFunction
object. init_ : estimator
The estimator that provides the initial predictions. Set via the
init
argument orloss.init_estimator
. estimators_ : ndarray of DecisionTreeRegressor,shape (n_estimators,
loss_.K
) The collection of fitted subestimators.
loss_.K
is 1 for binary classification, otherwise n_classes.
Notes
The features are always randomly permuted at each split. Therefore, the best found split may vary, even with the same training data and
max_features=n_features
, if the improvement of the criterion is identical for several splits enumerated during the search of the best split. To obtain a deterministic behaviour during fitting,random_state
has to be fixed.References
J. Friedman, Greedy Function Approximation: A Gradient Boosting Machine, The Annals of Statistics, Vol. 29, No. 5, 2001.
 Friedman, Stochastic Gradient Boosting, 1999
T. Hastie, R. Tibshirani and J. Friedman. Elements of Statistical Learning Ed. 2, Springer, 2009.
Methods
apply
(self, X)Apply trees in the ensemble to X, return leaf indices. decision_function
(self, X)Compute the decision function of X
.fit
(self, X, y[, sample_weight, monitor])Fit the gradient boosting model. get_params
(self[, deep])Get parameters for this estimator. predict
(self, X)Predict class for X. predict_log_proba
(self, X)Predict class logprobabilities for X. predict_proba
(self, X)Predict class probabilities for X. score
(self, X, y[, sample_weight])Returns the mean accuracy on the given test data and labels. set_params
(self, \*\*params)Set the parameters of this estimator. staged_decision_function
(self, X)Compute decision function of X
for each iteration.staged_predict
(self, X)Predict class at each stage for X. staged_predict_proba
(self, X)Predict class probabilities at each stage for X. 
__init__
(self, loss=’deviance’, learning_rate=0.1, n_estimators=100, subsample=1.0, criterion=’friedman_mse’, min_samples_split=2, min_samples_leaf=1, min_weight_fraction_leaf=0.0, max_depth=3, min_impurity_decrease=0.0, min_impurity_split=None, init=None, random_state=None, max_features=None, verbose=0, max_leaf_nodes=None, warm_start=False, presort=’auto’, validation_fraction=0.1, n_iter_no_change=None, tol=0.0001)[source]¶

apply
(self, X)[source]¶ Apply trees in the ensemble to X, return leaf indices.
New in version 0.17.
Parameters:  X : {arraylike, sparse matrix}, shape (n_samples, n_features)
The input samples. Internally, its dtype will be converted to
dtype=np.float32
. If a sparse matrix is provided, it will be converted to a sparsecsr_matrix
.
Returns:  X_leaves : arraylike, shape (n_samples, n_estimators, n_classes)
For each datapoint x in X and for each tree in the ensemble, return the index of the leaf x ends up in each estimator. In the case of binary classification n_classes is 1.

decision_function
(self, X)[source]¶ Compute the decision function of
X
.Parameters:  X : {arraylike, sparse matrix}, shape (n_samples, n_features)
The input samples. Internally, it will be converted to
dtype=np.float32
and if a sparse matrix is provided to a sparsecsr_matrix
.
Returns:  score : array, shape (n_samples, n_classes) or (n_samples,)
The decision function of the input samples, which corresponds to the raw values predicted from the trees of the ensemble . The order of the classes corresponds to that in the attribute classes_. Regression and binary classification produce an array of shape [n_samples].

feature_importances_
¶  Return the feature importances (the higher, the more important the
 feature).
Returns:  feature_importances_ : array, shape (n_features,)
The values of this array sum to 1, unless all trees are single node trees consisting of only the root node, in which case it will be an array of zeros.

fit
(self, X, y, sample_weight=None, monitor=None)[source]¶ Fit the gradient boosting model.
Parameters:  X : {arraylike, sparse matrix}, shape (n_samples, n_features)
The input samples. Internally, it will be converted to
dtype=np.float32
and if a sparse matrix is provided to a sparsecsr_matrix
. y : arraylike, shape (n_samples,)
Target values (strings or integers in classification, real numbers in regression) For classification, labels must correspond to classes.
 sample_weight : arraylike, shape (n_samples,) or None
Sample weights. If None, then samples are equally weighted. Splits that would create child nodes with net zero or negative weight are ignored while searching for a split in each node. In the case of classification, splits are also ignored if they would result in any single class carrying a negative weight in either child node.
 monitor : callable, optional
The monitor is called after each iteration with the current iteration, a reference to the estimator and the local variables of
_fit_stages
as keyword argumentscallable(i, self, locals())
. If the callable returnsTrue
the fitting procedure is stopped. The monitor can be used for various things such as computing heldout estimates, early stopping, model introspect, and snapshoting.
Returns:  self : object

get_params
(self, deep=True)[source]¶ Get parameters for this estimator.
Parameters:  deep : boolean, optional
If True, will return the parameters for this estimator and contained subobjects that are estimators.
Returns:  params : mapping of string to any
Parameter names mapped to their values.

predict
(self, X)[source]¶ Predict class for X.
Parameters:  X : {arraylike, sparse matrix}, shape (n_samples, n_features)
The input samples. Internally, it will be converted to
dtype=np.float32
and if a sparse matrix is provided to a sparsecsr_matrix
.
Returns:  y : array, shape (n_samples,)
The predicted values.

predict_log_proba
(self, X)[source]¶ Predict class logprobabilities for X.
Parameters:  X : {arraylike, sparse matrix}, shape (n_samples, n_features)
The input samples. Internally, it will be converted to
dtype=np.float32
and if a sparse matrix is provided to a sparsecsr_matrix
.
Returns:  p : array, shape (n_samples, n_classes)
The class logprobabilities of the input samples. The order of the classes corresponds to that in the attribute classes_.
Raises:  AttributeError
If the
loss
does not support probabilities.

predict_proba
(self, X)[source]¶ Predict class probabilities for X.
Parameters:  X : {arraylike, sparse matrix}, shape (n_samples, n_features)
The input samples. Internally, it will be converted to
dtype=np.float32
and if a sparse matrix is provided to a sparsecsr_matrix
.
Returns:  p : array, shape (n_samples, n_classes)
The class probabilities of the input samples. The order of the classes corresponds to that in the attribute classes_.
Raises:  AttributeError
If the
loss
does not support probabilities.

score
(self, X, y, sample_weight=None)[source]¶ Returns the mean accuracy on the given test data and labels.
In multilabel classification, this is the subset accuracy which is a harsh metric since you require for each sample that each label set be correctly predicted.
Parameters:  X : arraylike, shape = (n_samples, n_features)
Test samples.
 y : arraylike, shape = (n_samples) or (n_samples, n_outputs)
True labels for X.
 sample_weight : arraylike, shape = [n_samples], optional
Sample weights.
Returns:  score : float
Mean accuracy of self.predict(X) wrt. y.

set_params
(self, **params)[source]¶ Set the parameters of this estimator.
The method works on simple estimators as well as on nested objects (such as pipelines). The latter have parameters of the form
<component>__<parameter>
so that it’s possible to update each component of a nested object.Returns:  self

staged_decision_function
(self, X)[source]¶ Compute decision function of
X
for each iteration.This method allows monitoring (i.e. determine error on testing set) after each stage.
Parameters:  X : {arraylike, sparse matrix}, shape (n_samples, n_features)
The input samples. Internally, it will be converted to
dtype=np.float32
and if a sparse matrix is provided to a sparsecsr_matrix
.
Returns:  score : generator of array, shape (n_samples, k)
The decision function of the input samples, which corresponds to the raw values predicted from the trees of the ensemble . The classes corresponds to that in the attribute classes_. Regression and binary classification are special cases with
k == 1
, otherwisek==n_classes
.

staged_predict
(self, X)[source]¶ Predict class at each stage for X.
This method allows monitoring (i.e. determine error on testing set) after each stage.
Parameters:  X : {arraylike, sparse matrix}, shape (n_samples, n_features)
The input samples. Internally, it will be converted to
dtype=np.float32
and if a sparse matrix is provided to a sparsecsr_matrix
.
Returns:  y : generator of array of shape (n_samples,)
The predicted value of the input samples.

staged_predict_proba
(self, X)[source]¶ Predict class probabilities at each stage for X.
This method allows monitoring (i.e. determine error on testing set) after each stage.
Parameters:  X : {arraylike, sparse matrix}, shape (n_samples, n_features)
The input samples. Internally, it will be converted to
dtype=np.float32
and if a sparse matrix is provided to a sparsecsr_matrix
.
Returns:  y : generator of array of shape (n_samples,)
The predicted value of the input samples.