AI Terminology Course
AI Terminology
/
Beginner

Features

Definition

The individual, measurable properties or characteristics of a phenomenon being observed that a machine learning model uses to make its predictions.

Explain Like I'm New

The 'clues' the AI uses to solve the puzzle. If you want an AI to predict the price of a house, the 'Features' are the number of bedrooms, the square footage, the zip code, and the age of the house.

Real World Example

In an AI built to predict if an email is spam, the Features might be: 'Contains the word FREE', 'Sent at 3:00 AM', 'Has multiple exclamation points!!!', and 'Sender address matches the recipient'.

Common Use Cases

  • •Data engineering
  • •Feature extraction

Interview Questions

basic

  • If an AI is predicting the weather, what might be a good 'Feature' to give it?

intermediate

  • What is 'Feature Engineering'?

Flash Cards

Question

Weather feature?

Click to reveal answer
Answer

Current temperature, humidity percentage, barometric pressure, or historical cloud cover data.

Question

What is Feature Engineering?

Click to reveal answer
Answer

The process of human data scientists manually selecting, modifying, or creating new features from raw data to make the machine learning algorithm work better (e.g., combining 'latitude' and 'longitude' into a single 'distance_from_city_center' feature).