Page 26 - CT_AI_Class-6

 Social Network analysis: Finding groups or communities of people with similar interests or
connections. It helps in understanding relationships and improving user recommendations.

Unlabelled Input data Model Analysis Clustering Clusters / Groupings
(Detects Hidden Patterns)

The model explores the data
and finds similarities and The model forms groups (clusters)
relationships. Similar footballs are grouped based on patterns and similarities,
together into clusters.
without using any labels.

The diagram shows unsupervised learning using different footballs as unlabelled data. The model
studies the footballs and finds similarities such as colour, size and design. Without any given labels,
it learns patterns on its own. It then groups similar footballs into clusters. When new footballs are
added, the model places them into the most suitable group based on learned patterns.

Reinforcement Learning
Reinforcement Learning (RL) is a type of machine learning in which systems learn through
trial and error. Instead of being given correct answers, the machine explores different actions

and improves its performance by receiving rewards for correct decisions and penalties for
incorrect ones.

Over time, it focuses on actions that give the highest rewards and develops better strategies.
This approach is commonly used in areas like robotics, automation and smart decision-making
systems.
For example, an online recommendation system learns by suggesting different products or videos
to users without knowing their exact preferences initially. When users click or like a suggestion,

it receives a reward, while ignoring the suggestion acts as a penalty. Over time, it improves its
recommendations by learning from user responses.

ethical minds

AI systems, which rely on algorithmic thinking to process data and make decisions, can be biased if
they are trained on biased data.

Some applications of reinforcement learning are as follows:
 Game-playing AI: AI systems learn to play games like chess, Go or video games by trying
different moves and strategies. They receive rewards for winning or making good moves and
penalties for mistakes, which helps them improve over time.

24 Artificial Intelligence (CT & AI)-VI