Icon

Synthetic Data Generator Example on Titanic Data

This workflows generates synthetic numeric and nominal columns based on the distribution and statistics of the original columns and optionally the relationships between columns in the original data. It either replaces the original data with synthetic values (Example 1), or augments the original data by repeatedly generating synthetic data from it. Furthermore, it evaluates the quality of the data generation by performing an ANOVA test on the synthetic vs. original data.

Example 1: Anonymize data, keep known dependencies Compare means and variances in original/generated samples Example 2: Augment data Example workflow demonstrating the data generator components My workflow Read titanicdataNode 3Node 669Node 670Node 671Only age, sex, fareNode 677Node 678age (Gaussian)fare (Gamma distribution)sexage (Gaussian)10 000 rowsfare (Gamma distribution)10 000 rowssex10 000 rowsNode 698Node 699age (Gaussian)Node 702CSV Reader Number To String Table Manipulator One-way ANOVA Concatenate ConstantValue Column ConstantValue Column CSV Reader Joiner Joiner Synthetic DataGenerator (Numeric) Synthetic DataGenerator (Numeric) Synthetic DataGenerator (Nominal) Synthetic DataGenerator (Numeric) Synthetic DataGenerator (Numeric) Synthetic DataGenerator (Nominal) Row Filter CSV Reader Number To String Synthetic DataGenerator (Numeric) Column Filter Example 1: Anonymize data, keep known dependencies Compare means and variances in original/generated samples Example 2: Augment data Example workflow demonstrating the data generator components My workflow Read titanicdataNode 3Node 669Node 670Node 671Only age, sex, fareNode 677Node 678age (Gaussian)fare (Gamma distribution)sexage (Gaussian)10 000 rowsfare (Gamma distribution)10 000 rowssex10 000 rowsNode 698Node 699age (Gaussian)Node 702CSV Reader Number To String Table Manipulator One-way ANOVA Concatenate ConstantValue Column ConstantValue Column CSV Reader Joiner Joiner Synthetic DataGenerator (Numeric) Synthetic DataGenerator (Numeric) Synthetic DataGenerator (Nominal) Synthetic DataGenerator (Numeric) Synthetic DataGenerator (Numeric) Synthetic DataGenerator (Nominal) Row Filter CSV Reader Number To String Synthetic DataGenerator (Numeric) Column Filter

Nodes

Extensions

Links