[Diagram: Agent -> Sensors -> Percepts -> Actions -> Actuators -> Environment]
Key Points:
- Acts only on the basis of current perception.
- Ignores the rest of percept history.
- Based on If-Then rules.
- Environment should be fully observable.
Example: If temperature > 45°C, switch on AC (if fully observable).
Model-Based Reflex Agent:
- Works by finding the rule whose condition matches the current situation.
- Can work in a partially observable environment and track the situation.
- Agent keeps track of internal state, which is adjusted by each percept and depends on percept history.
- Model: Knowledge about how things happen in the world.
1) How the world is evolving.
2) How agent’s actions affect the world.
[Diagram: Model-based agent diagram]
, Goal-Based Agent:
- Focuses only on reaching the goal set.
- Agent takes decisions based on how far it is currently from the goal state.
- Every action is taken to minimize distance to the goal state.
- More flexible agent.
Key Points:
- Acts only on the basis of current perception.
- Ignores the rest of percept history.
- Based on If-Then rules.
- Environment should be fully observable.
Example: If temperature > 45°C, switch on AC (if fully observable).
Model-Based Reflex Agent:
- Works by finding the rule whose condition matches the current situation.
- Can work in a partially observable environment and track the situation.
- Agent keeps track of internal state, which is adjusted by each percept and depends on percept history.
- Model: Knowledge about how things happen in the world.
1) How the world is evolving.
2) How agent’s actions affect the world.
[Diagram: Model-based agent diagram]
, Goal-Based Agent:
- Focuses only on reaching the goal set.
- Agent takes decisions based on how far it is currently from the goal state.
- Every action is taken to minimize distance to the goal state.
- More flexible agent.
