Machine Learning with SAS Viya

Chapter 4: Preparing Your Data: Extract Features

Introduction

Extract Features

Model Studio: The Feature Extraction Node

Principal Component Analysis (PCA)

Robust Principal Analysis

Autoencoders

Handling Extreme or Unusual Values

Transforming Inputs: Mathematical Functions

Transforming Inputs: Binning

Model Studio: The “Best” Transformation

Demo 4.1: Transforming Inputs

Feature Selection

Details: The Curse of Dimensionality

Feature Selection Strategies

Model Studio: Feature Selection

Demo 4.2: Selecting Features

Demo 4.3: Saving a Pipeline to the Exchange

Variable Clustering

Demo 4.4: Clustering Inputs for Data Reduction

Quiz

Introduction

Often, the initial (structured) data set of raw features might be too large and unwieldy to be effectively ...