University of Florida
         Department of Electrical and Computer Engineering
                              Fall 1997
       EEL 5840: ELEMENTS OF MACHINE INTELLIGENCE

Instructors: Antonio A. Arroyo and Keith L. Doty
Class Web Site: http://www.mil.ufl.edu/eel5840
Text     Readings from the literature and class notes provided during the semester.

General Objectives
Provide an in-depth look at Machine Learning, both classical and modern, with a view toward grounding the
concepts in physical reality; to implement Machine Learning Algorithms in autonomous robots and contrast with
simulations of intelligent autonomous robots; to provide an "engineering approach" to the emerging field of
Machine Intelligence (MI); to impart a conceptual foundation on the principles behind current MI technology; to
provide greater understanding of the limitations of current learning theories.

Course Objectives
You will read a diverse technical literature on learning and intelligence and will attempt to apply different theories
and paradigms to physical autonomous robots, or to simulations of autonomous robots, in your homework
assignments. You will contrast your simulation studies with physical embodied studies, ascertaining the value and
limitations of each approach. In your project you will focus on a particular machine learning technique in depth and
develop an implementation on a real robot or a simulation. You will write a formal report on your project. You will
also present and demonstrate your project to the class.

Project
Your project should realize a particular machine learning theory on a real robot. The theory may be an existing one,
a modified existing theory, or one you have invented yourself.

You may also write a robot simulator to illustrate learning theories. However, you will need to work with someone
with a robot and try to apply your simulation results on a real machine. A number of simulators already exist in this
area. If you choose to simulate, you should explore what has already been done.

You may work in teams of two on a project, but project responsibility must be clearly delineated.

Robots for Homework and Projects
If you took EEL5666, you can use the robot you built in that course to satisfy homework and project requirements.
You can also purchase a robot kit or an assembled robot.

Grading
There will be two exams.

         Midterm Exam                30%
         Final Exam                  30%
         Homework and Project        40%
                        Total       100%




                                                                                                                    1
EEL 5840 Schedule
                   HOMEWORK ASSIGNMENTS: FALL SEMESTER 1997
WEEK              HOME SET   DUE READING
 1. 8/25          None                             Classical AI and Machine Intelligence
 2. 9/1                                    9/5     Classifiers
                  Holiday 9/1
 3. 9/8           Home Set 1               9/12    Neural Net Perspective
 4. 9/15          Home Set 2               9/19    Psychological Perspective on Learning and Intelligence:
                                                   Conditioned Response, Habituation…
 5. 9/22          Home Set 3               9/26    Adaptive Behavior and Ethology
 6. 9/29          Home Set 4               10/3    Reactive Behavior: Case Study
 7. 10/6          Home Set 5               10/10   Behavior with Memory : Case Study
 8. 10/13         Exam 1                   10/17   Biology based neural control: Neurophysiology
 9. 10/20         Home Set 6               10/24   Non-linear Dynamics Approach to Robot Behavior
10. 10/27         Home Set 7                       Reinforcement Learning, Markov Decision Processes, ,
                                                   Gradient Search
11. 11/3          Homecoming             11/7-8    Dynamic Programming, Q-Learning
12. 11/10         Holiday 11/11                    Q-Learning,
13. 11/17         Home Set 9               11/21   Evolutionary Learning: Genetic Algorithms
14. 11/24         Holiday 11/27-28                 Complexity Theory: Order and Chaos
15. 12/1                                           Artificial Life
16. 12/8          Final Exam       12/15           Presentation of Projects



References
We will read from many of the references cited, as time permits. This list will be increased as the semester
progresses.
[Beer 1990] Randall Beer, Intelligence as Adaptive Behavior. Academic Press, N.Y., 1990.
[Beer, Quinn, et al 1997] R. Beer, R.D. Quinn, H.J. Chiel, and R.E. Ritzmann. Biologically Inspired Approaches to
    Robotics. Communications of the ACM, Vol. 40, No. 3, March 1997, pp. 31-38.
[Braitenberg 1984] V. Braitenberg. Vehicles, Experiments in Synthetic Psychology. MIT Press, Cambridge, MA,
    1984.
[Denardo 1982] E.V. Denardo, Dynamic Programming: Models and Applications. Prentice Hall, Englewood Cliffs,
    N.J.
[Gage 1993] Douglas W. Gage. Randomized Search Strategies with Imperfect Sensors. Proceedings of SPIE Mobile
    Robots VIII, Vol. 2058, Boston, Sept.9-10, 1993, pp. 270-279.
[Gleick 1987] J. Gleick, Chaos: Making a New Science. N.Y., Viking, 1987.
[Goldberg 1989] D. Goldberg. Genetic Algorithms in Search, Optimization, and Machine Learning. Addison-
    Wesley, Reading, MA., 1989.
[Holland 1975] J.H. Holland, Adaptation in Natural and Artificial Systems. University of Michigan Press, Ann
    Arbor, 1975.
[Kaufman 1993] S. Kaufman, The Origins of Order: Self-Organization and Selection in Evolution. N.Y., Oxford
    University Press, 1993.
[Lorenz 1973] K. Lorenz. Foundations of Ethology. Springer-Verlag, N.Y.,1973.
[Maes 1991] P. Maes, editor. Designing Autonomous Agents. MIT Press, Cambridge, MA, 1991.
[Mataric` 1994] M.J. Mataric`. Interaction and Intelligent Behavior. Ph.D. Thesis, Dept. of EECS, MIT, Cambridge,
    MA, 1994.
[Moffat and Frijda 1994] David Moffat and Nico H. Frijda. Where there’s a Will there’s an Agent. ECAI-94
    Workshop on Agent Theories, Architectures & Languages. Springer-Verlag, Amsterdam, The Netherlands, Aug.
    1994, pp. 244-260.
[Resnick 1994] M. Resnick. Turtles, Termites, and Traffic Jams. MIT Press, Cambridge, MA., 1994.
[Reynolds 1987] Craig W. Reynolds. Flocks, Herds, and Schools: A Distributed Behavioral Model. Proceedings of
    SIGGRAPH ’87 (Computer Graphics 21(4), July1987, edited by Maureen C. Stone, pp. 25-34).
[Sutton 1988] Richard S. Sutton. Learning to Predict by the Methods of Temporal Differences. Machine Learning,
    3:9-44, Kluwer Academic Publishers, Boston, 1988.
[Walter 1950] W. G. Walter. An Imitation of Life. Scientific American, May 1950, pp. 42-45.
[Whitehead and Lin 1995] Steven D. Whitehead and Long-Ji Lin. Review of Reinforcement Learning. Elsevier
    Science B.V., 1995.

                                                                                                               2

syl_f97.doc

  • 1.
    University of Florida Department of Electrical and Computer Engineering Fall 1997 EEL 5840: ELEMENTS OF MACHINE INTELLIGENCE Instructors: Antonio A. Arroyo and Keith L. Doty Class Web Site: http://www.mil.ufl.edu/eel5840 Text Readings from the literature and class notes provided during the semester. General Objectives Provide an in-depth look at Machine Learning, both classical and modern, with a view toward grounding the concepts in physical reality; to implement Machine Learning Algorithms in autonomous robots and contrast with simulations of intelligent autonomous robots; to provide an "engineering approach" to the emerging field of Machine Intelligence (MI); to impart a conceptual foundation on the principles behind current MI technology; to provide greater understanding of the limitations of current learning theories. Course Objectives You will read a diverse technical literature on learning and intelligence and will attempt to apply different theories and paradigms to physical autonomous robots, or to simulations of autonomous robots, in your homework assignments. You will contrast your simulation studies with physical embodied studies, ascertaining the value and limitations of each approach. In your project you will focus on a particular machine learning technique in depth and develop an implementation on a real robot or a simulation. You will write a formal report on your project. You will also present and demonstrate your project to the class. Project Your project should realize a particular machine learning theory on a real robot. The theory may be an existing one, a modified existing theory, or one you have invented yourself. You may also write a robot simulator to illustrate learning theories. However, you will need to work with someone with a robot and try to apply your simulation results on a real machine. A number of simulators already exist in this area. If you choose to simulate, you should explore what has already been done. You may work in teams of two on a project, but project responsibility must be clearly delineated. Robots for Homework and Projects If you took EEL5666, you can use the robot you built in that course to satisfy homework and project requirements. You can also purchase a robot kit or an assembled robot. Grading There will be two exams. Midterm Exam 30% Final Exam 30% Homework and Project 40% Total 100% 1
  • 2.
    EEL 5840 Schedule HOMEWORK ASSIGNMENTS: FALL SEMESTER 1997 WEEK HOME SET DUE READING 1. 8/25 None Classical AI and Machine Intelligence 2. 9/1 9/5 Classifiers Holiday 9/1 3. 9/8 Home Set 1 9/12 Neural Net Perspective 4. 9/15 Home Set 2 9/19 Psychological Perspective on Learning and Intelligence: Conditioned Response, Habituation… 5. 9/22 Home Set 3 9/26 Adaptive Behavior and Ethology 6. 9/29 Home Set 4 10/3 Reactive Behavior: Case Study 7. 10/6 Home Set 5 10/10 Behavior with Memory : Case Study 8. 10/13 Exam 1 10/17 Biology based neural control: Neurophysiology 9. 10/20 Home Set 6 10/24 Non-linear Dynamics Approach to Robot Behavior 10. 10/27 Home Set 7 Reinforcement Learning, Markov Decision Processes, , Gradient Search 11. 11/3 Homecoming 11/7-8 Dynamic Programming, Q-Learning 12. 11/10 Holiday 11/11 Q-Learning, 13. 11/17 Home Set 9 11/21 Evolutionary Learning: Genetic Algorithms 14. 11/24 Holiday 11/27-28 Complexity Theory: Order and Chaos 15. 12/1 Artificial Life 16. 12/8 Final Exam 12/15 Presentation of Projects References We will read from many of the references cited, as time permits. This list will be increased as the semester progresses. [Beer 1990] Randall Beer, Intelligence as Adaptive Behavior. Academic Press, N.Y., 1990. [Beer, Quinn, et al 1997] R. Beer, R.D. Quinn, H.J. Chiel, and R.E. Ritzmann. Biologically Inspired Approaches to Robotics. Communications of the ACM, Vol. 40, No. 3, March 1997, pp. 31-38. [Braitenberg 1984] V. Braitenberg. Vehicles, Experiments in Synthetic Psychology. MIT Press, Cambridge, MA, 1984. [Denardo 1982] E.V. Denardo, Dynamic Programming: Models and Applications. Prentice Hall, Englewood Cliffs, N.J. [Gage 1993] Douglas W. Gage. Randomized Search Strategies with Imperfect Sensors. Proceedings of SPIE Mobile Robots VIII, Vol. 2058, Boston, Sept.9-10, 1993, pp. 270-279. [Gleick 1987] J. Gleick, Chaos: Making a New Science. N.Y., Viking, 1987. [Goldberg 1989] D. Goldberg. Genetic Algorithms in Search, Optimization, and Machine Learning. Addison- Wesley, Reading, MA., 1989. [Holland 1975] J.H. Holland, Adaptation in Natural and Artificial Systems. University of Michigan Press, Ann Arbor, 1975. [Kaufman 1993] S. Kaufman, The Origins of Order: Self-Organization and Selection in Evolution. N.Y., Oxford University Press, 1993. [Lorenz 1973] K. Lorenz. Foundations of Ethology. Springer-Verlag, N.Y.,1973. [Maes 1991] P. Maes, editor. Designing Autonomous Agents. MIT Press, Cambridge, MA, 1991. [Mataric` 1994] M.J. Mataric`. Interaction and Intelligent Behavior. Ph.D. Thesis, Dept. of EECS, MIT, Cambridge, MA, 1994. [Moffat and Frijda 1994] David Moffat and Nico H. Frijda. Where there’s a Will there’s an Agent. ECAI-94 Workshop on Agent Theories, Architectures & Languages. Springer-Verlag, Amsterdam, The Netherlands, Aug. 1994, pp. 244-260. [Resnick 1994] M. Resnick. Turtles, Termites, and Traffic Jams. MIT Press, Cambridge, MA., 1994. [Reynolds 1987] Craig W. Reynolds. Flocks, Herds, and Schools: A Distributed Behavioral Model. Proceedings of SIGGRAPH ’87 (Computer Graphics 21(4), July1987, edited by Maureen C. Stone, pp. 25-34). [Sutton 1988] Richard S. Sutton. Learning to Predict by the Methods of Temporal Differences. Machine Learning, 3:9-44, Kluwer Academic Publishers, Boston, 1988. [Walter 1950] W. G. Walter. An Imitation of Life. Scientific American, May 1950, pp. 42-45. [Whitehead and Lin 1995] Steven D. Whitehead and Long-Ji Lin. Review of Reinforcement Learning. Elsevier Science B.V., 1995. 2