The second part of the course
Introduction to AI Using the functional Lisp programming
language, we will continue to study
some basic concepts and techniques of
artificial intelligence.
TDDC65 Artificial intelligence and Lisp A series of lectures and laboratory
Peter Dalenius exercises.
petda@ida.liu.se Russell & Norvig (2003) Artificial
Department of Computer and Information Science
Linköping University Intelligence: A Modern Approach
Final written exam December 20, 2008.
What is intelligence? What is intelligence?
Intelligence is a very general mental capability that,
among other things, involves the ability to
– reason, It is only a word that people use to
– plan, name those unknown processes with
– solve problems,
– think abstractly,
which our brains solve problems we call
– comprehend complex ideas, hard.
– learn quickly and learn from experience.
It is not merely book learning, a narrow academic
skill, or test-taking smarts. Rather, it reflects a
broader and deeper capability for comprehending our
surroundings – “catching on”, “making sense” of
things, or “figuring out” what to do.
Gottfredson, Linda (1997) Mainstream Science on Intelligence. Intelligence 24:1 Quote attributed to Marvin Minsky, one of the first researchers in AI.
What is artificial intelligence? Different approaches to AI
Artificalintelligence tries to understand Which concept What do we measure
is most important? success against?
how we think in order to build intelligent
entities (e.g. machines or computer Humans Rationality
programs).
Thought
Systems that Systems that
AI is a new science, only 50 years old, think like humans think rationally
processes
but is heavily influenced by other fields.
Systems that act Systems that act
Behaviour like humans rationally
1
1. Systems that think like humans 2. Systems that act like humans
But how do humans think?!
Can machines pass
Can machines
a behavioral
Computer models Experimental methods think?
from artificial intelligens from psychology intelligence test?
Cognitive science The Turing test
Trying to build models and theories
about how humans perceive and think.
Turing, Alan (1950) Computing Machinery and Intelligence
The Turing test (one variety) Skills needed to pass the test
Part 2: A is replaced
Part 1: Two persons. by a computer program.
Natural
language processing
Knowledge representation
A B A B
Automated reasoning
Machine learning
? ?
Goal: Try to guess which one is male/female.
A helps the interrogator to make the wrong
decision, B helps making the right decision. We will look at some of these in the course…
3. Systems that think rationally 4. Systems that act rationally
Aristotle’s laws of thought: Syllogisms Rational agents
– Something or someone that acts (a machine or a
Formalized patterns of reasoning: Logic computer program)
Any solveable problem expressed in logic can – Autonomous, perceiving the environment,
persisting over time, adapting to change, pursuing
be solved! goals
Problems with this approach: Advantages of this approach:
– Difficult to express informal knowledge using logic – More general than logic approach. There is more
to rationality than just correct reasoning.
– Computational explosion when trying to draw – A completely rational agent has all the skills
conclusions from a large knowledge base needed to pass a Turing test.
– Easier to develop systems when you can define a
degree of rationality for your specific project.
2
Influence from other disciplines Intelligent agents
An agent perceives its environment through sensors and
Philosophy execute actions through its actuators. The input at any given
time is called
Mathematics percept, and the
Economics complete history
Neuroscience of all percepts is
called a percept
Psychology sequence. The agent
Computer engineering is controlled by the
agent program,
Control theory and cybernetics which implements
Linguistics a function from
input to output.
Intelligent agents What is rational?
Agent Environment
Performance measure
Sensors – Predefined way of telling how good I am.
Percepts
Prior knowledge of the environment
Agent – What do I know about the world?
program Set of possible actions
– What can I do?
Actuators Percept sequence
Actions
– What has happened so far?
Specifying the task environment Properties of task environments
Agent Performance Environment Actuators Sensors
measure Fully or partially observable
Taxi driver Safe, fast, Roads, other Steering, Camera, Deterministic or stochastic
legal, traffic, accelerator, speedometer,
comfortable customers brake, signal GPS Episodic or sequential
Part-picking Percentage of Conveyor belt Jointed arm Camera, joint
robot parts in correct with parts, bins and hand angle sensors Static or dynamic
bin
Interactive Maximize Set of Display Keyboard Discrete or continuous
tutor student’s score students, exercises, entry
on test testing agency suggestions
Single agent or multiagent
PEAS = Performance, Environment, Actuators, Sensors
3
Agent architectures Simple reflex agent
Agent Sensors
Simple reflex agent
Model-based reflex agent What the world
Environment
is like now
Goal-based agent
Utility-based agent
Learning agents
What action I
Condition-action rules should do now
Actuators
Model-based reflex agent Goal-based agent
Agent Sensors Agent Sensors
State State
What the world What the world
Environment
Environment
How the world evolves is like now How the world evolves is like now
What my actions do What my actions do
What it will be like
if i do action A
What action I What action I
Condition-action rules should do now Goals should do now
Actuators Actuators
Utility-based agent Learning agent Performance standard
Agent Sensors Agent Sensors
State Critic
What the world
Environment
Environment
How the world evolves is like now feedback
changes
What my actions do Learning Performance
What it will be like
if i do action A element element
knowledge
How happy I will be learning
Utility in such a state goals
What action I Problem
should do now generator
Actuators Actuators
4
WITAS UAV Project Architecture
