liveBook · Manning

Feature association

As explained in Outlier Detection in Python

A feature association is a statistical relationship or dependence between feature values across observations.

Feature association is the dependence between feature values across observations. Synthetic data can preserve it by generating features sequentially, predicting each later feature from already populated features rather than sampling every column independently. section 11.2.2

Baseball feature synthesis

A baseball dataset generates numeric features from histograms with jitter, then uses regression or classification models to generate later columns from earlier synthetic columns.

How Outlier Detection in Python covers it

How deep it goes
Compares independent, fully sequential, and limited-prior generation, but does not develop a general theory of dependence.
Examples
Demonstrates association using baseball features including At_bats and RBIs, comparing independent and left-to-right synthetic data.
What it assumes
Assumes familiarity with tabular features and distributions, while demonstrating the generation logic from individual columns onward.
What you can do after
Apply it via NumPy, pandas, and scikit-learn to generate synthetic columns while controlling which prior features supply predictions.
How it connects
Connects feature association with marginal distributions, joint distributions, histograms, kernel density estimation, and predictive models.
Cover of Outlier Detection in Python
liveBookmanning.com
Read Outlier Detection in Python on liveBook Get the book at manning.com

This explainer was generated by Manning's knowledge graph from the text of Outlier Detection in Python. The section links point to the parts of the book it draws on.