Introduction to Artificial Intelligence Concepts

Course Information

• Course title: BPSY121-4: Artificial Intelligence
• Instructor: Dr. Suganthi J
• University: CHRIST Deemed to be University, Bangalore, India

Artificial Intelligence (AI)

• Definition (Artificial): Made by humans, especially in imitation of natural processes.
• Definition (AI in the context of the course): Artificial intelligence leverages computers and machines to mimic the problem-solving and decision-making capabilities of the human mind.
• Definition (Intelligence): The ability to acquire and apply knowledge and skills.

What is Artificial Intelligence?

• A method of making computers, computer-controlled robots, or software think intelligently like the human mind.

How Psychologists Define Intelligence

• Ability to learn
• Ability to recognize problems
• Ability to solve problems

Intelligent Agents

• Agents in Artificial Intelligence
• An agent is anything that perceives its environment through sensors and acts upon that environment through actuators.
• An agent operates in a cycle of perceiving, thinking, and acting.
• An intelligent agent is an autonomous entity that acts on an environment using sensors and actuators to achieve goals.
• Examples of agents:
• Human Agent: uses senses (eyes, ears) and actuators (hands, legs, voice).
• Robotic Agent: utilizes cameras, infrared, NLP for sensors and motors for actuators.
• Software Agent: leverages keystrokes, file contents as sensory inputs and displays outputs on a screen.

Sensors and Actuators

• Sensor: A device that detects changes in the environment and sends information to other electronic devices. An agent observes its environment through sensors.
• Actuator: A component of machines that converts energy into motion to move and control a system. Examples include electric motors, gears, and rails. An actuator is the mechanism that allows the "effector" to execute an action. Actuators typically include electric motors, hydraulic or pneumatic cylinders.

Effectors

• Devices that affect the environment. Examples include legs, wheels, arms, fingers, wings, fins, and display screens.

Agent and Environment

• An agent is anything that can be viewed as perceiving its environment through sensors and acting upon that environment through actuators.
• An agent's behavior is described by the agent function that maps any given percept sequence to an action.

Agent Terminology

• Performance Measure of Agent: The criteria used to evaluate how successful an agent is.
• Behavior of Agent: The action an agent performs after a given sequence of percepts.
• Percept: An agent's perceptual input at a given instance.
• Percept Sequence: The history of all that an agent has perceived until a particular point in time.
• Agent Function: A map from the percept sequence to an action.

Good Behavior: The Concept of Rationality

• Rational Agents: Agents that use logic or reason.
• Systems that can reasonably be called intelligent: systems that can act rationally.
• Rationality: judged by the following points:
• Performance measure that defines the criteria of success.
• Agent's prior knowledge of the environment.
• The actions that the agent can perform.
• Agent's percept sequence to date.

Four Main Rules of an AI Agent

• Rule 1: An AI agent must have the ability to perceive the environment.
• Rule 2: The observation must be used to make decisions.
• Rule 3: Decision should result in an action.
• Rule 4: The action taken by an AI agent must be a rational action.

Omniscience, Learning and Autonomy

• Omniscience: An omniscient agent knows the actual outcome of its actions and can act accordingly.
• Learning: A rational agent that gathers information and learns from what it perceives. Initial configuration may reflect prior knowledge of the environment.
• Autonomy: A rational agent should be autonomous—it should learn what it can to compensate for partial or incorrect prior knowledge.

Environment

• Everything around the agent, but not part of the agent itself.
• The environment provides the agent with something to sense and act on.
• Environments are mostly considered to be non-feministic.

Nature of Environments

• Fully observable vs. Partially observable
• Static vs. Dynamic
• Discrete vs. Continuous
• Deterministic vs. Stochastic
• Single-agent vs. Multi-agent
• Episodic vs. Sequential
• Known vs. Unknown
• Accessible vs. Inaccessible

Structure of an AI Agent

• Agent = Architecture + Agent Program
• Architecture: The machinery that the agent executes on, such as a device with sensors and actuators. Example: robotic car, camera, PC.
• Agent Program: An implementation of an agent function.
• Agent Function: A map from the percept sequence (history of all that an agent has perceived to date) to an action.

Types of Agents

• Simple reflex agents
• Model-based reflex agents
• Goal-based agents
• Utility-based agents

Learning Agents

• Type of agent that can learn from its experiences.
• Starts with basic knowledge and then adapts automatically through learning.
• Learning Element: Responsible for making improvements by learning from the environment.
• Critic: Provides feedback on how well the agent is performing against a fixed standard.
• Performance Element: Responsible for selecting external actions.
• Problem Generator: Suggests actions to lead to new and informative experiences.

PEAS Representation

• A model to define AI agent properties.
• Made up of four words:
• P: Performance measure
• E: Environment
• A: Actuators
• S: Sensors

PEAS Examples (Self-Driving Car)

• Performance: Safety, time, legal driving, comfort.
• Environment: Roads, other vehicles, road signs, pedestrians.
• Actuators: Steering, accelerator, brake, signal, horn.
• Sensors: Camera, GPS, speedometer, odometer, accelerometer, sonar.

Example Agents with PEAS

• Medical Diagnosis: Performance: healthy patient, minimized cost; Environment: patient, hospital, staff; Actuators: tests, treatments; Sensors: keyboard (entry of symptoms).
• Vacuum Cleaner: Performance: cleanness, efficiency, battery life, security; Environment: room, table, wood floor, carpet, various obstacles; Actuators: wheels, brushes, vacuum extractor; Sensors: camera, dirt detection sensor, cliff sensor, bump sensor, infrared wall sensor, joint angle sensors.
• Part-Picking Robot: Performance: percentage of parts in correct bins; Environment: conveyor belt w various obstacles, bins; Actuators: jointed arms, hands; Sensors: camera, infrared wall sensor, joint angle sensors.

Problem Solving

• Problem Solving Agents: Decide what to do by finding a sequence of actions that leads to a desirable state or solution.
• Requires thinking through a series of moves to a goal state.
• Search: The computation used by problem-solving agents.

Problem Solving Process

• Goal Formulation: The initial phase, defining the desired target/goal that requires action.
• Problem Formulation: Determining the steps required to reach the goal.
• Search: Finding a sequence of actions to achieve the goal.
• Execution: Implementing the recommended actions one by one.

Problems and Solutions

• Problem definition (formal):
• Initial State
• Actions
• Transition Model
• Goal Test
• Path Cost

Example Problems (Standardized/Toy)

• Vacuum world problem: robot moves around a grid of cells sucking dirt.
• Grid world problem: a two-dimensional rectangular array of cells where agents move around.

Example Problem (8 Puzzle)

• A 3 x 3 grid with 8 numbered tiles and one blank space
• The goal is to reach a specified goal state