Feature Engineering

> (Statistics|Probability|Machine Learning|Data Mining|Data and Knowledge Discovery|Pattern Recognition|Data Science|Data Analysis)

Table of Contents

1 - About

In Feature engineering, you are:

  • creating derived features
  • normalize them

3 - Example

3.1 - Fraud detection

Data Mining - Fraud Detection

  • Aggregated variables. Example: aggregated transaction count per account in last 24 hours to spot abnormal amount
  • Mismatch variables. Example: delivery country <> normal delivery country
  • Risk tables. Probability Risks grouped by country, state, IP Address, etc..
Advertising

4 - Documentation / Reference

data_mining/feature_engineering.txt · Last modified: 2018/11/11 15:23 by gerardnico