Artificial Intelligence (AI) · Beginner level

Beginner level · Module 2: How AI Learns from Data · Lesson 1 of 3

Data, Features and Labels

See how examples with features and labels teach an AI system to recognise patterns.

In this lesson you will

  • Identify features and labels in a data set
  • Explain why data quantity and quality matter

Key concepts

training datafeatureslabelsdata quality

The lesson

Machine learning systems learn from examples. To recognise ripe mangoes, a system is shown many photos (the features) each marked 'ripe' or 'unripe' (the labels).

The system adjusts itself to find patterns that connect features to labels. With more varied, correctly labelled examples, it learns patterns that work on new cases.

Poor data produces poor AI: wrong labels, too few examples or examples from only one situation — only mangoes on white plates — teach the wrong patterns.

Step by step

  1. Collect 20 photos of two kinds of object.
  2. Label them carefully.
  3. Identify what varies in your photos.
  4. Predict where the model might fail.

Practice activity

Prepare a small labelled data set for the no-code classifier.

Expected result: A labelled set with notes on its variety and gaps.

Check your understanding

  1. What is a label?

    Show answer

    The correct answer attached to a training example.

  2. Why does variety in training data matter?

    Show answer

    So the system learns patterns that work in new situations.

Taught live by an instructor — online or in person. Free course lessons are for self-study and do not lead to a certificate — see registration options.