H2O PCA Compute

KNIME H2O Machine Learning Integration version 3.7.0.v201811291237 by KNIME AG, Zurich, Switzerland

Compute a Principal Component Analysis (PCA) model using H2O. This model can later be used to reduce the dimensionality of a dataset using the PCA Apply node.

Options

General Settings

Column selection

Select columns used for model training.

Ignore constant columns

Select to ignore constant columns.

Use static random seed

Select to use static seed for randomization.

Algorithm Settings

Transformation method

Specify the transformation method for the training data (None, Standardize, Normalize, Demean, or Descale) (transform).

PCA method

Specify the algorithm to use for computing the principal components (pca_method).

Rank of matrix approximation

Specify the rank of matrix approximation (the dimension of the principal components). Note that the dimension cannot be larger than the dimension of the input data (k).

Number of max iterations

Specify the number of training iterations (max_iterations).

Use all factor levels

Specify whether to use all factor levels in the possible set of predictors (use_all_factor_levels).

Impute missing values

Specify whether to impute missing entries with the column mean value (impute_missing). Note that if this option is set to false, any rows with missing values will be removed.

Max runtime in secs

Maximum allowed runtime in seconds for model training (max_runtime_secs).

Input Ports

H2O Frame with the input data.

Output Ports

H2O Principal Component Analysis model.

Best Friends (Incoming)

• Table to H2O (50 %) Manipulator
• Parameter Optimization Loop Start (50 %) LoopStart

Best Friends (Outgoing)

• H2O PCA Apply (100 %) Manipulator
• H2O Model to MOJO (< 1 %) Manipulator

Workflows

• 01_Guided_Analytics_for_ML_Automation

Installation

To use this node in KNIME, install KNIME H2O Machine Learning Integration from the following update site: