Intelligent Decision Support Systems

1
Intelligent Decision Support Systems

• IE 8541 / HUMF 8001

2
Decision Support Systems

• Computer tool that helps people make better
  decisions, faster
• Leaves the human in control of decision making.

3
Examples of Situations in which DSSs may be used

• Help medical personnel in a hospital setting to
  consider various diagnoses and treatment plans,
• Help engineers to analyze, develop and compare
  solution options,
• Help military planners to generate alternative
  plans of military maneuvers,
• Help search and rescue workers to prioritize
  which of many critical activities to attend next.

4
The need for DSSs in the 21rst Century

• The move towards global comepetion in a global
  market place has increased the market for more
  complex systems and products,
• At the same time companies and organizations must
  produce these systems and products faster, and
  with less personnel,
• DSSs provide a way to help manage this
  complexity,

5
Example of How a DSS can multiply human
capabilities

• NASA Mars Rover program demonstrated that a DSSs
  can enable people to effectively manage more
  complex systems than would be possible with out
  DSS assistance.
• Operations staff at JPL could produce 40 more
  activities with the rovers when using a DSS, than
  not.

6
Intelligent Decision Support Systems

• Intelligent DSSs perform some part of the problem
  solving for the user.
• Non-intelligent DSSs my simply structure or
  present knowledge in a way that facilitates
  problem solving or decision making.
• This course will emphasize intelligent DSSs which
  assist in complex problem solving situations,
  usually requiring expertise

7
A Continuum of Human Control and Automation
Agent assisted by human
Design environment
Autonomous system
Intelligent assistant
Pencil and paper
Critic
Adapted from Fischer et al. Critics, and
Emerging Approach to knowledge-based ,
International Journal of Man-Machine Studies, 35
(5) 695-721, 1991.
8
Challenges

• DSSs are hard to characterize as points on this
  scale
• They may be highly automated in some portions,
  and human driven in others.
• A single system may incorporate multiple styles
  of interaction.
• Need additional ways in which to characterize
  DSSs design space, map out specific planning
  functions.

9
Course goals

• Students will apply information learned in
  readings
• To develop a decision support system on a topic
  of their own expertise or choosing (often from
  the students work or research).
• The DSS does not have to be super high tech or
  use advanced programming techniques some very
  successful systems have been implemented in
  Excel).

10
Assignments

• A semester project
• Summaries of readings. A one page synopsis of
  each article (or chapter) assigned will be due at
  the beginning of each class. Additional
  exercises may be assigned as well.
• Presentation of an article to the class
• In-class exercises
• Midterm
• Project progress report
• Quiz
• Final project report
• In-class presentation on the results of your
  project

11
Grade Breakdown

• Summaries 10 percent
• Class participation 3 percent
• Project Proposal 10 percent
• Midterm 10 percent
• Project Progress Report 8 percent
• Quiz 7 percent
• Article presentation 7 percent
• Final project presentation 10 percent
• Project Final Report 35 percent

12
Human Factors Minor

• This class can be used towards the Human Factors
  minor (Masters or PhD.)
• Human Factors minor program requirements
  http//education.umn.edu/kin/kinesiology/HFminor.h
  tml
• If IE students want to use this course towards
  the minor, they must use the HF 8001 course
  number.

13
Revised HF Minor Requirements coming soon

• New requirements for the HF program will be
  coming out in February, pending the provosts
  approval.
• While I cannot post the new requirements until
  they are approved, I am permitted to inform
  students of what the new requirements are likely
  to be, upon request.

14
For Next Thursday 1/18/07

• Read Development of Expertise by J. R.
  Anderson, prepare summary.
• For next Tuesday
• Read Verbal Reports as Data,