ENCS3340 - Artificial Intelligence
Introduction
What is Artificial Artificial Intelligence?
• Let’s define intelligence first.
• Intelligence, based on Wikipedia, is the capacity for:
• Abstraction
• Logic
• Understanding
• Self-awareness
• Learning
• Emotional knowledge
• Reasoning
• Planning
• Creativity
• Critical thinking
• Problem-solving.
I.e. the ability to perceive or infer information, and to retain it as knowledge to
be applied towards adaptive behaviors within an environment or context.
1
What is Artificial Artificial Intelligence?
• No standard definition of AI among those working in the field.
• AI has even been defined as:
“... the collection of problems and methodologies studied by artificial
intelligence researchers." - Luger and StubbleField
• The textbook discusses four different schools of thought:
• Machines that think like humans
• Machines that act like humans
• Machines that think rationally
• Machines that act rationally
2
2 Dimensions to Describe AI
• 4 points of views: Systems that can ….
Cognitive Engineering /
Approach Rational Approach
Behavior act like humans act rationally
Reasoning think like humans think rationally
• Engineering Approach:
• Tries to find optimal solution.
• No matter how (not necessarily what human do)
• Cognitive Approach:
• Tries to understand the process
• Tries to reproduce human behavior (even if wrong results)
3
A Test for Intelligence
The Turing Test
• Proposed by Alan Turing in 1950
• If a human interrogator cannot tell the computer and human
apart, then the computer is intelligent
• Measures the intelligence of a computer vs. a human
• Turing predicted that by 2000, a machine might have a 30% chance of fooling
a person for 5 minutes
4
The Turing Test
What capabilities would a computer need to have to pass the Turing Test?
• Natural language processing
• Knowledge representation
• Automated reasoning
• Machine learning
• ….
5
Turing Test: Criticism
• Not reproducible
• Not constructive
• Not a good way to solve practical problems
• Not amenable to mathematical analysis
• Machine intelligence designed w.r.t. humans
• Some human behavior is not intelligent
• Some intelligent behavior may not be human
Intelleginet
Unintelligent Turing behavior
human Test humans do
behavior not do
Human Intelligent
behavior behavior 6
What do we do in AI?
• Search / Heuristic Search
• Automatic Game Playing
• Planning Systems
• Knowledge Representation
• Logic and Inference
• Natural Language Processing
• Dealing with uncertainty – probability and decision theory
• Machine Learning
• Vision & Robotics
• Philosophical Issues
• …
7
Disciplines involved in AI
• Computer science:
• Algorithms and data structures to implement theories
• Philosophy:
• logic, methods of reasoning, foundation of learning, rationality
• Mathematics:
• formal representation and proof, algorithms, computations, (un)decidability, (in)tractability,
probability
• Psychology:
• Experimental techniques (psycholinguistics, ...)
• Linguistics:
• Knowledge representation, grammars
• mechanical engineering.
• ….
8
Applications of AI
Games
Chess Go
Deep Blue, 1997 AlphaGo, 2017
AlphaZero, 2017
9
Applications of AI
Robotics
Mars rovers RoboCup
10
Applications of AI
Autonomous Driving
11
Applications of AI
Natural Language Understanding
Automatic speech recognition Question Answering
Machine Translation 12
Applications of AI
Computer Vision
Scene Understanding Activity Recognition
Face Detection
13
Applications of AI
And Many More Applications …..
• Recommendation Systems • Semantic Web
• Search Engines • Reasoning
• Spam Filtering • Medical Diagnosis Systems
• Planning • ….
• Navigation Systems
14
Course Topics
• Search
• Uninformed search [tree/graph traversal]
• Informed search, heuristics [A*]
• Constraint satisfaction problems [map coloring]
• Games [Chess]
• Logic [Propositional and First-Order Logic, inference/proving theorems]
15
Course Topics
• Machine Learning [learning from examples]
• Supervised and unsupervised learning
• Decision trees
• Naïve Bayes
• Neural Networks
• Clustering
• Applications [if time permits]
• Natural Language Processing
• Computer Vision
• Robotics
• …
16
Intelligent Agents
Recap: What is AI?
• 4 points of views: Systems that can ….
Cognitive Engineering /
Approach Rational Approach
Behavior act like humans act rationally
Reasoning think like humans think rationally
• The textbook advocates "acting rationally”
• Acting rationally: Rational Agent
• Rational behavior: doing the right thing
• The right thing: that which is expected to maximize goal achievement, given the
available information
• Doesn't necessarily involve thinking – e.g., blinking reflex – but thinking should
be in the service of rational action
18
Agents
• An agent is anything that can be viewed as perceiving its environment
through sensors and acting upon that environment through actuators
• Abstractly, an agent is a function from percept histories to actions
[f: P* → A]
• The agent program runs on the physical architecture to produce f
19
Example of Agents
• Human agent:
• eyes, ears, and other organs for sensors
• hands, legs, mouth, and other body parts for actuators
• Robotic agent:
• cameras and infrared range finders for sensors
• various motors for actuators
• software agent:
• functions as sensors: information provided as input to functions in the form of
encoded bit strings or symbols
• functions as actuators: results/outputs of the function
20
Example: Vacuum-cleaner world
• Percepts: location and contents, e.g.,
[A,Dirty]
• Actions: Left, Right, Suck, NoOp
21
Rational Agents
• An agent should strive to "do the right thing", based on what it can perceive
and the actions it can perform.
• The right action is the one that will cause the agent to be most successful
• Performance measure: An objective criterion for success of an agent's
behavior
• Possible performance measures of a vacuum-cleaner agent:
• amount of dirt cleaned up,
• amount of time taken,
• amount of electricity consumed,
• amount of noise generated, etc
• Rational Agent: For each possible percept sequence, a rational agent should
select an action that is expected to maximize its performance measure,
given the evidence provided by the percept sequence and whatever built-in
knowledge the agent has.
22
PEAS
• PEAS: Performance measure, Environment, Actuators, Sensors
• Performance measure: used to evaluate how well an agent solves the task at
hand
• Environment: surroundings beyond the control of the agent
• Actuators: determine the actions the agent can perform
• Sensors: provide information about the current state of the environment
23
PEAS examples
Agent: automated taxi driver
• Performance measure: Safe, fast, legal, comfortable trip, maximize profits
• Environment: Roads, other traffic, pedestrians, customers
• Actuators: Steering wheel, accelerator, brake, signal, horn
• Sensors: Cameras, sonar, speedometer, GPS, odometer, engine sensors,
keyboard
24
PEAS examples
Agent: Medical diagnosis system
• Performance measure: Healthy patient, minimize costs, lawsuits
• Environment: Patient, hospital, staff
• Actuators: Screen display (questions, tests, diagnoses, treatments, referrals)
• Sensors: Keyboard (entry of symptoms, findings, patient's answers)
25
PEAS examples
Agent: Part-picking robot in a factory
• Performance measure: Percentage of parts in correct bins
• Environment: Conveyor belt with parts, bins
• Actuators: Jointed arm and hand
• Sensors: Camera, joint angle sensors
26
Environment Types
• Fully observable (vs. partially observable):
• An agent's sensors give it access to the complete state of the environment at
each point in time.
• Deterministic (vs. stochastic):
• The next state of the environment is completely determined by the current state
and the action executed by the agent.
• Episodic (vs. sequential):
• The agent's experience is divided into atomic "episodes" (each episode consists
of the agent perceiving and then performing a single action), and the choice of
action in each episode depends only on the episode itself.
27
Environment Types
• Static (vs. dynamic):
• The environment is unchanged while an agent is deliberating. (The environment
is semidynamic if the environment itself does not change with the passage of
time but the agent's performance score does)
• Discrete (vs. continuous):
• A limited number of distinct, clearly defined percepts and actions.
• Single agent (vs. multiagent):
• An agent operating by itself in an environment.
28
Environment Types
• The environment type largely determines the agent design
• The real world is (of course) partially observable, stochastic, sequential, dynamic,
continuous, multi-agent
29
Agent types
Five basic types in order of increasing generality:
• Simple reflex agents
• select actions on the basis of the current percept, ignoring the rest of the percept
history.
• Model-based reflex agents
• maintain some sort of internal state that depends on the percept history and thereby
reflects at least some of the unobserved aspects of the current state.
• Goal-based agents
• needs some sort of goal information that describes situations that are desirable
• Utility-based agents
• chooses actions to maximize its utility
• Learning agents
30