sklearn.linear_model.SGDRegressor 파라미터

sklearn.linear_model.SGDRegressor 파라미터

class sklearn.linear_model.SGDRegressor(loss='squared_error', *, penalty='l2', alpha=0.0001, l1_ratio=0.15, fit_intercept=True, max_iter=1000, tol=0.001, shuffle=True, verbose=0, epsilon=0.1, random_state=None, learning_rate='invscaling', eta0=0.01, power_t=0.25, early_stopping=False, validation_fraction=0.1, n_iter_no_change=5, warm_start=False, average=False)

 

sklearn.linear_model.SGDRegressor 파라미터

 

lossstr, default=’squared_error’

penalty{‘l2’, ‘l1’, ‘elasticnet’}, default=’l2’

alphafloat, default=0.0001

l1_ratiofloat, default=0.15

fit_interceptbool, default=True

max_iterint, default=1000

tolfloat, default=1e-3

 

shufflebool, default=True

verboseint, default=0

epsilonfloat, default=0.1

random_stateint, RandomState instance, default=None

learning_ratestr, default=’invscaling’

eta0float, default=0.01

power_tfloat, default=0.25

early_stoppingbool, default=False

 

validation_fractionfloat, default=0.1

 

n_iter_no_changeint, default=5

warm_startbool, default=False

averagebool or int, default=False