What is the difference between supervised and unsupervised machine learning?

The core difference between these two types of machine learning lies in the data used to train them:

• Supervised Learning uses a "labeled" dataset. This means the input data comes with the correct answers or categories already attached. The model learns to map inputs to the known outputs, essentially learning by example under the guidance of a "teacher."
• Unsupervised Learning uses an "unlabeled" dataset. The model is given data without any explicit instructions or target answers. It must independently look for patterns, structures, or groupings within the data itself.

Real-World Examples

Supervised Learning: Email Spam Filtering An email provider uses a vast database of emails that have already been manually tagged as "Spam" or "Not Spam" by users. The model learns the characteristics of spam (such as specific keywords or sender patterns) so that when a new, incoming email arrives, it can accurately label it as spam or safe.

Unsupervised Learning: Customer Segmentation A retail company takes a large list of its customer data—such as purchase history, age, and location—but doesn't tell the model what to look for. The model identifies natural clusters, such as finding a specific group of shoppers who tend to buy outdoor gear on weekends. The company can then use these discovered groupings to create more relevant marketing campaigns.