1. Chapter 1: Computational Intelligence and Knowledge
What is Computational Intelligence (CI)? (1.1)
Agents in the world (1.2)
Representation & Reasoning (1.3)
Examples of Applications Domains (1.4)
Our Approach to Teaching CI
D. Poole, A. Mackworth, and R. Goebel, Computational Intelligence: A Logical Approach,
Oxford University Press, January 1998

2. What is Computational Intelligence? (1.1)
CI = The study of the design of intelligent agents
An agent is something that acts in an environment (dogs, airplanes, humans…)
An intelligent agent is an agent that acts intelligently:
its actions are appropriate for its goals and circumstances
it is flexible to changing environments and goals (it adapts)
it learns from experience
it makes appropriate choices given perceptual limitations and finite computation
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

3. What is Computational Intelligence? (1.1) (cont.)
Artificial or Computational Intelligence?
The field is often called Artificial Intelligence (confusing term: artificial pearl is a fake pearl !).
Scientific goal: to understand the principles that make intelligent behavior possible, in natural or artificial systems.
Engineering goal: to specify methods (programs) for the design of useful, intelligent artifacts.
Analogy between studying flying machines and thinking machines (bats, birds vs. airplanes). In order to produce the functionality of intelligent behavior, it is not necessary to reproduce the structural connections of the human body (Icarus goal). Functional Equivalence  Structural Equivalence
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

4. Computational Intelligence researchers are interested in testing general hypotheses about the nature of intelligence.
They do so by building machines that are intelligent and which does not necessarily mimic humans or group of agents.
Can Computers really think ? [Turing] versus Can airplanes really fly ?
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

5. What is Computational Intelligence? (1.1) (cont.)
Central hypotheses of CI
Symbol-system hypothesis:
Reasoning is symbol manipulation.
Church–Turing thesis:
Any symbol manipulation can be carried out on a Turing machine
Turing Machine = idealization of a digital computer with an unbounded amount of memory.
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

6. Agents in the world (1.2)
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

7. Representation and Reasoning (1.3)
To use these inputs an agent needs a representation of them.
 Knowledge
Most common sense tasks rely on a lot of knowledge.
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

8. Representation and Reasoning (1.3) (cont.)
Representation and Reasoning Systems
Problem  representation  computation
A representation and reasoning system (RRS) consists of
Language to communicate with the computer.
A way to assign meaning to the symbols.
Procedures to compute answers or solve problems.
Example of RRS’s:
Programming languages: Fortran, C++,…
Natural Language
**** We want something between these extremes.
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

9. Applications using RRS (1.4)
The Autonomous delivery robot roams around an office environment and delivers coffee, parcels,…
The Diagnostic assistant helps a human troubleshoot problems and suggests repairs or treatments. e.g., electrical problems, medical diagnosis.
The Infobot searches for information on a computer system or network.
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

10. CS 426 Artificial Intelligence, Ch
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

11. Applications (1.4) (cont.)
Autonomous Delivery Robot
Example inputs:
Prior knowledge: knowledge about its capabilities, objects it may encounter, info about its environment such as maps.
Past experience: which actions are useful and when, what objects are there, how its actions affect its position.
Goals: what it needs to deliver and when, tradeoffs between acting quickly and acting safely.
Observations: about its environment from cameras, sonar, sound, laser range finders, or keyboards.
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

12. Applications (1.4) (cont.)
What does the Delivery Robot need to do?
Determine where “Craig’s” office is. Where coffee is…
Find a path between locations.
Plan how to carry out multiple tasks.
Make default assumptions about where “Craig” is.
Make tradeoffs under uncertainty: should it go near the stairs?
Learn from experience.
Sense the world, avoid obstacles, pickup and put down coffee.
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

13. CS 426 Artificial Intelligence, Ch
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

14. Applications (1.4) (cont.)
Diagnostic Assistant
Example inputs:
Prior knowledge: how switches and lights work, how malfunctions manifest themselves, what information tests provide, the side effects of repairs.
Past experience: the effects of repairs or treatments, the prevalence of faults or diseases.
Goals: fixing the device and tradeoffs between fixing or replacing different components, patient live longer (medical diagnosis).
Observations: symptoms of a device or a patient.
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

15. Applications (1.4) (cont.)
Subtasks for the diagnostic assistant
Derive the effects of faults and interventions.
Search through the space of possible fault or diseases.
Explain its reasoning to the human who is using it.
Derive possible causes for symptoms; rule out other causes.
Plan courses of tests and treatments to address the problems.
Reason about the uncertainties/ambiguities given symptoms.
Trade off alternate courses of action.
Learn about what symptoms are associated with the faults,
the effects of treatments, and the accuracy of tests.
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

16. Applications (1.4) (cont.)
Infobot
Infobot interacts with an information environment:
It takes in high-level, perhaps informal, queries.
It finds relevant information.
It presents the information in a meaningful way.
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

17. Applications (1.4) (cont.)
Infobot inputs
Prior knowledge: the meaning of words, the types of information sources, and how to access information.
Past experience: where information can be obtained, the relative speed of various servers, and information about the preferences of the user.
Goals: the information it needs to find out; tradeoffs between the volume and quality of information and the expense involved.
Observations: what information is at the current sites; what links are available; the load on various connections.
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

18. Applications (1.4) (cont.)
Example subtasks for the Infobot
Derive information that is only implicit in a knowledge base.
Interact in natural language.
Find good representations of knowledge.
Explain how an answer was derived and why some information was unavailable.
Make conclusions about lack of knowledge, and determine conflicting knowledge.
Use default reasoning about where to obtain different information.
Make tradeoffs between cheap but unreliable information sources and more expensive but more comprehensive [cost vs. quality]
Learn about what knowledge is available where, and what information the user is interested in [user preferences].
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

19. Applications (1.4) (cont.)
Common Tasks of the Three Domains
1. Modeling the environment: The robot needs to be able to model the physical
environment, its own capabilities, and the mechanics of delivering parcels.
For the diagnostic assistant, it should know how treatments work, or for the
infobot, it should know how the information can be obtained.
2. Evidential reasoning or perception: Given some observations (such as
symptoms) about the world, determine what is really in the world (disease)
(system identification, or diagnosis).
3. Action: Given a model of the world and a goal, determine what should be done
in term of actions (robot roves, treatments and tests).
4. Learning from past experiences: Learn about the specific case and the
population of cases (how the robot sensors actually work?, know how well
particular diseases responds to particular treatments? the particular patient
background.
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra

20. Our approach to teaching CI
Our goal is to study these four tasks.
We build the tools needed from the bottom up.
We start with some restrictive simplifying assumptions and lift them as we get more sophisticated representations and more powerful reasoning strategies.
The theory and practice are built from solid
foundations.
CS 426 Artificial Intelligence, Ch.1: Computational Intelligence & Knowledge
Dr. Djamel Bouchaffra