sklearn.manifold.TSNE 파라미터

TSNE

class sklearn.manifold.TSNE(n_components=2, *, perplexity=30.0, early_exaggeration=12.0, learning_rate='warn', n_iter=1000, n_iter_without_progress=300, min_grad_norm=1e-07, metric='euclidean', init='warn', verbose=0, random_state=None, method='barnes_hut', angle=0.5, n_jobs=None, square_distances='legacy')

 

TSNE 파라미터

n_componentsint, default=2

perplexityfloat, default=30.0

early_exaggerationfloat, default=12.0

learning_ratefloat or ‘auto’, default=200.0

n_iterint, default=1000

n_iter_without_progressint, default=300

min_grad_normfloat, default=1e-7

metricstr or callable, default=’euclidean’

init{‘random’, ‘pca’} or ndarray of shape (n_samples, n_components), default=’random’

verboseint, default=0

random_stateint, RandomState instance or None, default=None

methodstr, default=’barnes_hut’

 

anglefloat, default=0.5

n_jobsint, default=None

square_distancesTrue or ‘legacy’, default=’legacy’