skglm.LinearSVC

class skglm.LinearSVC(C=1.0, max_iter=50, max_epochs=50000, p0=10, verbose=0, tol=0.0001, fit_intercept=True, warm_start=False, ws_strategy='subdiff')

LinearSVC estimator, with hinge loss.

The optimization objective for LinearSVC is:

`C xx sum_(i=1)^(n_"samples") max(0, 1 - y_i beta^T X[i, :]) + 1/2 ||beta||_2 ^ 2`

i.e. hinge datafit loss (non-smooth) + l2 regularization (smooth)

To solve this, we solve the dual optimization problem to stay in our framework of smooth datafit and non-smooth penalty. The dual optimization problem of SVC is:

`1/2 ||(yX)^T w||_2 ^ 2 - \sum_(i=1)^(n_"samples") w_i + \sum_(i=1)^(n_"samples") [0 <= w_i <= C]`

The primal-dual relation is given by:

`w = \sum_(i=1)^(n_"samples") y_i w_i X[i, :]`

Parameters:

Cfloat, optional

Regularization parameter. The strength of the regularization is inversely proportional to C. Must be strictly positive.

max_iterint, optional

The maximum number of iterations (subproblem definitions).

max_epochsint

Maximum number of CD epochs on each subproblem.

p0int

First working set size.

verbosebool or int

Amount of verbosity.

tolfloat, optional

Stopping criterion for the optimization.

fit_interceptbool, optional (default=True)

Whether or not to fit an intercept.

warm_startbool, optional (default=False)

When set to True, reuse the solution of the previous call to fit as initialization, otherwise, just erase the previous solution.

ws_strategystr

The score used to build the working set. Can be fixpoint or subdiff.

Attributes:

coef_array, shape (n_features,)

parameter vector (`w` in the cost function formula)

sparse_coef_scipy.sparse matrix, shape (n_features, 1)

sparse_coef_ is a readonly property derived from coef_

intercept_float

constant term in decision function.

dual_array, shape (n_samples,)

dual of the solution.

n_iter_int

Number of subproblems solved to reach the specified tolerance.

__init__(C=1.0, max_iter=50, max_epochs=50000, p0=10, verbose=0, tol=0.0001, fit_intercept=True, warm_start=False, ws_strategy='subdiff')

Methods

__init__([C, max_iter, max_epochs, p0, ...])
decision_function(X)	Predict confidence scores for samples.
densify()	Convert coefficient matrix to dense array format.
fit(X, y)	Fit LinearSVC classifier.
get_metadata_routing()	Get metadata routing of this object.
get_params([deep])	Get parameters for this estimator.
predict(X)	Predict class labels for samples in X.
score(X, y[, sample_weight])	Return the mean accuracy on the given test data and labels.
set_params(**params)	Set the parameters of this estimator.
set_score_request(*[, sample_weight])	Request metadata passed to the score method.
sparsify()	Convert coefficient matrix to sparse format.