Skip to content
BENIBIX Learn

Supervised Learning

Also known as Learning from Labels

The model learns from data. Each row has the right answer on it.

Think of it like

Think of it like a teacher who marks your work. Each sum comes back right or wrong.

Example

Emma gives a model 50,000 old loans, each marked paid back or not paid back. It learns to guess which new loans will be paid back.

Why it matters

When you have good labels, this is usually the most accurate way to learn. Getting the labels is often the hard part.

Definition

Supervised learning is machine learning from examples that come with the right answers. Each example has features and a label. The model learns to guess the label from the features. It can sort things into groups or guess numbers. Most machine learning in business works this way.
Supervised learning is the first panel. Each dot comes with its colour, and that colour is its label.

Common mistake

Starting with too few labels. A few hundred examples is rarely enough for a hard task.

Key takeaways

  1. 1
    Labelled examples. Each one has the answer.
  2. 2
    Learns the link. From features to label.
  3. 3
    Most common type. The workhorse of business AI.

Compare

Learn it properly