Modules
Spark MLlib from Elixir, over Spark Connect.
Classifiers, and the models they fit.
Attributes of a BinaryLogisticRegressionSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a BinaryLogisticRegressionTrainingSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a BinaryRandomForestClassificationSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a BinaryRandomForestClassificationTrainingSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a fitted DecisionTreeClassificationModel.
Attributes of a fitted FMClassificationModel.
Attributes of a FMClassificationSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a FMClassificationTrainingSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a fitted GBTClassificationModel.
Attributes of a fitted LinearSVCModel.
Attributes of a LinearSVCSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a LinearSVCTrainingSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a fitted LogisticRegressionModel.
Attributes of a LogisticRegressionSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a LogisticRegressionTrainingSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a fitted MultilayerPerceptronClassificationModel.
Attributes of a MultilayerPerceptronClassificationSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a MultilayerPerceptronClassificationTrainingSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a fitted NaiveBayesModel.
Attributes of a fitted RandomForestClassificationModel.
Attributes of a RandomForestClassificationSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a RandomForestClassificationTrainingSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Clustering estimators, and the models they fit.
Attributes of a fitted BisectingKMeansModel.
Attributes of a BisectingKMeansSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a ClusteringSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a fitted DistributedLDAModel.
Attributes of a fitted GaussianMixtureModel.
Attributes of a GaussianMixtureSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a fitted KMeansModel.
Attributes of a KMeansSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a fitted LDAModel.
Attributes of a fitted LocalLDAModel.
K-fold cross-validation: fit every param map on every fold, and keep the best.
What a cross-validation found: the winning model, and the score of every param map.
An unfitted estimator: inert data, no session, no server-side anything.
Evaluators: one metric over a scored frame.
A metric, as inert data: what to measure and on which columns.
Frequent pattern mining.
Attributes of a fitted FPGrowthModel.
Feature transformers, and the models the fitted ones produce.
Attributes of a fitted BucketedRandomProjectionLSHModel.
Attributes of a fitted ChiSqSelectorModel.
Attributes of a fitted CountVectorizerModel.
Attributes of a fitted IDFModel.
Attributes of a fitted ImputerModel.
Attributes of a fitted MaxAbsScalerModel.
Attributes of a fitted MinHashLSHModel.
Attributes of a fitted MinMaxScalerModel.
Attributes of a fitted OneHotEncoderModel.
Attributes of a fitted PCAModel.
Attributes of a fitted RFormulaModel.
Attributes of a fitted RobustScalerModel.
Attributes of a fitted StandardScalerModel.
Attributes of a fitted StringIndexerModel.
Attributes of a fitted UnivariateFeatureSelectorModel.
Attributes of a fitted VarianceThresholdSelectorModel.
Attributes of a fitted VectorIndexerModel.
Attributes of a fitted Word2VecModel.
Converting between a Vector column and an ordinary array of numbers.
An operator that is neither fitted nor applied: its work is one method on the server's own helper object.
Spark's Vector and Matrix, both ways.
A fitted model: a reference into the session's ML cache, and the context needed to use it.
One operator the registry knows, as inert data.
One parameter an operator accepts.
A sequence of stages, fitted as one.
A fitted pipeline: its stages, in order, with every estimator replaced by the model it fitted.
ML commands and relations, built as protos.
Collaborative filtering.
Attributes of a fitted ALSModel.
Regressors, and the models they fit.
Attributes of a fitted AFTSurvivalRegressionModel.
Attributes of a fitted DecisionTreeRegressionModel.
Attributes of a fitted FMRegressionModel.
Attributes of a fitted GBTRegressionModel.
Attributes of a fitted GeneralizedLinearRegressionModel.
Attributes of a GeneralizedLinearRegressionSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a GeneralizedLinearRegressionTrainingSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a fitted IsotonicRegressionModel.
Attributes of a fitted LinearRegressionModel.
Attributes of a LinearRegressionSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a LinearRegressionTrainingSummary — what an estimator recorded while it fitted. Latu.ML.summary/1 is what hands you one.
Attributes of a fitted RandomForestRegressionModel.
Read what an MlCommand answered.
A Spark sparse vector, kept sparse.
Statistical tests that run where the data is: chi-square, correlation, Kolmogorov-Smirnov.
A model's training summary: a second server-side object, and the one the cache really drops.
One split instead of k folds: fit every param map once, and keep the best.
What a train/validation split found: the winning model, and one metric per param map.
A transformer that needs no fitting — VectorAssembler and its kind.