You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Provides a simple interface to Scikit-Learn's various matrix factorization models.
Since they provide a consistent training interface, you can simply choose the class name of the method you want to use and specify options relevant to that method in the options option. For available options, take a look at the table of parameters in the Scikit-Learn documentation for each class.
(required) Scikit-learn class to use to fit the matrix factorization. Should be the name of a class in the package sklearn.decomposition that has a fit_transform() method and a components attribute. Supported classes: NMF, SparsePCA, ProjectedGradientNMF, FastICA, FactorAnalysis, PCA, RandomizedPCA, LatentDirichletAllocation, TruncatedSVD
'NMF', 'SparsePCA', 'ProjectedGradientNMF', 'FastICA', 'FactorAnalysis', 'PCA', 'RandomizedPCA', 'LatentDirichletAllocation' or 'TruncatedSVD'
options
Options to pass into the constructor of the sklearn class, formatted as a JSON dictionary (potentially without the {}s). E.g.: 'n_components=200, solver="cd", tol=0.0001, max_iter=200'