Please enable JavaScript.
Coggle requires JavaScript to display documents.
Agents in Artificial Intelligence - Coggle Diagram
Agents in Artificial Intelligence
Definition
A computer program or system that is created to monitor its environment, make decisions, and take actions in order to accomplish a certain goal or set of goals is known as an agent.
Agent architectures
An AI agent uses architecture as its execution platform.
Percept
Perceptual inputs
Perceptual Sequence
The decision is based on part of the perceptual sequence.
Agent function
Agent programs implement agent functionality. On the physical architecture, an agent program creates function f.
Omniscience
Omniscient knows the actual outcome
Information gathering
Exploration
Learning
Agent program
Agent programs carry out agent functions. An agent software program executes on the physical architecture to produce function f.
Task Environment
PEAS Descriptors
Performance
Environment
Actuators
Sensors
Features/Properties of Environment from the point of view of an agent
Fully Observable vs Partially Observable
Deterministic vs Stochastic
Competitive vs Collaborative
Single-agent vs Multi-agent
Static vs Dynamic
Discrete vs Continuous
Episodic vs Sequential
Known vs Unknown
Vacuum cleaner world
Percepts. Eg. location and status
Actions. Eg. Move left, move right, suck dirt, do nothing
Agent = Architecture + Program
Simple reflex agent
Model-based reflex agent
Goal-based agent
Utility-based agent
Human agent, Robotic agent, software agent
Software sensors and actuators
Software sensors, Ex: Virtual GPS Sensor
Software actuators. Ex: Virtual Motor Actuator
Mixed Examples: Virtual Traffic Light
Types of agents in AI
Deliberative agents
Hybrid agents
Goal-based agents
Utility-based agents
Multi-agent systems
Rational agents
Definition
characteristics
A rational agent does the “right thing”.
Types of rational agents in AI (e.g. reflex, model-based, goal-based)
Performance measure
Defines the success criterion
Agent prior knowledge of its environment
Best possible actions that an agent can perform
The sequence of percepts.
The sequence of percepts.
possesses the capacity to learn from one's past experiences and acts or makes decisions based on those learnings.
obtains fundamental knowledge from the past and automatically acts and adapts using that learning.
has four thematic elements.
Learning element
Critic
Performance element
Performance element