Introduction

Agent: an entity that has goals or preferences and tries to perform a series of actions that yeild the best/optimal expected outcome given these goals.

Percept sequence: the complete history of content the agent has perceived.

Environment: the given instantiation of the agent.

Environment type:

• Fully vs Partially observable

• Deterministic vs. Stochastic

  The outcome of an action is completely determined by the state of the environment and the action taken by an agent.

• Episodic vs. Sequential

  Episodic environment can be divided into episodes which have no influence on each other

  Short-term actions do not have long-term consequences

• Static vs. Dynamic

  Environment doesn’t change while the agent is deliberating

• Discrete vs. Continuous

  The distinct states of the environmement is finite/infinite

Playing soccer: Partially observable, stochastic, dynamic, sequential, continous

Shopping for used books on the internet: Partially observable, stochastic, dynamic, episodic sequential

World: an environmant and the agents that reside within it create a world.

Simple reflex agent: one that doesn’t think about the consequences of its actions, but rather selects an action based solely on the current state of the world. Example: A basic thermostat used in a home to control the heating or cooling system.

Model-based reflex agent: maintains some internal state which depends on the percept history, useful if the current environment cannot be fully described by the current percept. Example: An autonomous robot vacuum cleaner recording the environment and navigates through the room.

Utility-based agents: compares the desirability of different environment states via a utility function. This allows the comparison of different goal states and action sequences and tradeoffs between different goals. Example: ADC.

Goal-based agents: makes decisions in order to achieve a set of predefined goals, in addition to maintaining internal state.

Problem-solving agents:

Single-state problem formulation: complete observable

A problem is defined by four items: (1) Initial state. (2) Actions(or successor function $S(x)$). (3) Goal test. (4) Path cost (additive).

A solution is a sequence of actions leading from initial state to goal state.