Page 25 - CT_AI_Class-6

In this data, weight refers to how heavy the fruit is and size describes its shape or form, such as
medium, long, round or small. These are called features.
The model is trained using this labelled data, where each combination of weight and size is linked
to a specific fruit type. By learning from these examples, the model begins to understand patterns
between the features and the correct labels.

Once trained, if the model is given new data, such as weight 120 grams and size long, it can
predict that the fruit is a banana. This shows how labelled data helps the model learn and make
accurate predictions.

Labelled Data
Prediction Output

AI Model Apple
Apple Banana
Grapes

Orange Grapes Banana

Orange

Test Data

Unsupervised learning
Unsupervised learning is a type of machine learning in which machines study data to identify

hidden patterns, structures or groupings. In this approach, the model learns from the patterns in
data without any predefined labels or outputs.

Unlike supervised learning, unsupervised learning does not provide definite guidance or correct
answers to the model. Instead, the model independently explores the data to discover meaningful
patterns and relationships. It is commonly used for tasks such as clustering similar data points
and detecting anomalies in datasets.
Some applications of unsupervised learning are as follows:

 Customer segmentation: Grouping customers based on similar buying habits, preferences or
behaviour. It helps businesses target the right audience with suitable products and offers.

 Market Basket analysis: Identifying products that are frequently bought together by
customers. This helps stores in product placement and making better recommendations.
 Document clustering: Grouping similar articles or documents based on their content or topics.
It helps in organising large amounts of information efficiently.

 Anomaly detection: Detecting unusual patterns or activities that do not match normal
behaviour. It is useful in identifying fraud or suspicious transactions.

Introduction to AI & Everyday Examples 23