COSC 6368 --- Artificial Intelligence Spring 2006 ( Dr. Eick )

• The teaching of the course will be somewhat similar to the version of the course that was offered in Fall 2004. However, there will be more coverage in 2006 of evolutionary computing, reinforcement learning, and there will be less coverage of PROLOG, knowledge-based systems, belief networks and planning. Moreover, the teaching of the course will be more project-oriented: solutions of assignments and homeworks will be discussed in more detail in the lecture.
• This webpage is an evolving document. Most information still refers to the Fall 2004 teaching of the course and will be updated during the course of the semester.

Material Covered in COSC 6368

Course Organization

• I Introduction AI and Course Information (1 class)
• II Heuristic Search (4 classes)
• III Evolutionary Computing (2 classes)
• IV Machine Learning and Data Mining (5.5 classes)
• V FOPL, Logical Reasoning, and Resolution (3 classes)
• VI Reasoning in Uncertain Environments --- Naive Bayesian Systems and Belief Networks (3 classes)
• VII Ontologies, the Sematic Web and Knowledge-based Systems (1.5 classes)
• VIII Philosophical Foundations of AI (0.5 classes)
• Discussion of Course Projects and Homeworks (2 classes)
• Other Activities: two exams (2 classes), review (1 class).

Prerequisites

Students are expected to have the following background:

• Knowledge of basic computer science principles and basic programming skills. Ability to understand and analyze fairly complicated algorithms and data structures. The course will assume familiarity with such concepts as: big-O notation, queues and stacks, trees, graphs, etc.
• Familiarity with the basic concepts of discrete probability theory. However, it suffices to read and understand Chapter 13 of Russell and Norvig.
• Familiarity with the basic search techniques, e.g. breadth-first search that are usually covered in a data structure course. However, it more than enough to read and understand Chapter 3 of Russell and Norvig.
• It is assumed that you have some basis knowledge of propositional and predicate logic. It suffices to read and understand Chapters 7.3-7.5 and 8.1,8.2 of Russell and Norvig. Moreover, one course lecture will review the basic concept of predicate logic.
• Students should have basic programmig skills. Some assignments require programming; however, students can choose the programming language they like the most. It is not necessary to take COSC 4350 prior to taking COSC 6368.

The prerequisites for the class are important, but only up to a point. The real prerequisite for this course is the ability to solve abstract problems, to understand nontrivial algorithms, and to have basic programming and system development skills. To some people these skills come by more easily, whereas others get them by taking the corresponding classes. If you feel that you have these skills, you can easily make up for the prerequisites on your own.

Course Assignment Topics and Assignments

Due Dates and Exam Dates 2006

Class Transparencies

• 2006 AI Introduction Transparencies (will be used in class) and Brief History of AI.
• 2004 Search Transparencies:
  • Search1 (2 transparencies added at the end on Sept. 29, 2004 that discuss the Ungraded Search Homework Problem), Search2, Search3, Search4, Search5 (Games); Kamil on Backtracking and Mazes.
  • Rook+King vs. King Animation; R+K vs. K --- Problem Specification; Computer Bridge and Solving the Rook+King vs. King Endgame Problem --- a Case Study on Techniques to Solve Challenging Heuristic Search Problems; Results of the 2004 ACBL Computer Bridge Championships.
• 2004 Teaching Material Evolutionary Computing (EC):
  • Introduction to Evolutionary Computing
  • How to Use Evolutionary Computing to Solve "Real Problems".
  • To get a better understanding of the main ideas of EC read Eiben/Smith's Introduction to Evolutionary Algorithms; if you are interested in a more indepth review of EC research read Baeck's EC Survey Paper.
• 2004 Logical Reasoning Transparencies:
  • Russel Intro to Logic Transparencies ( will not be used in lecture with the exception of 17-19, 21-22, 26, 29, 30; only sections 6.3 and 6.4 of the textbook will be covered in class)
  • Russel Intro to Predicate Calculus Transparencies (transpencies 1-13 will be used in lecture)
  • Russel Resolution Transparencies (to be used in lecture).
  • Dr. Eick's Introduction to Resolution (contains answers for the Sept. 23 FOPL as a language exam), and simple examples of resolution proofs (Asci) --- the resolution method will be discussed in detail in class using a lot of examples.
  • Dr. Eick's additional Transparencies Predicate Logic and Resolution(Postscript, only transparencies 2 and 4 will be covered in 2004).
  • David A. Plaisted's Propositional Approaches to First Order Logic Theorem Proving (gives a good introduction on what researchers in the area of logical reasoning do)
• 2004 Machine Learning Transparencies, centering on Decision Trees, Neural Networks, Reinforcement Learning, and Data Mining:
  • Decision Trees
    • Teaching Material Machine Learning and Decision Trees by Tom Mitchell (CMU)
      • Tom Mitchell's "Introduction to Machine Learning" (transparencies 1-8 and 15-16 will be used)
      • Tom Mitchell's Decision Tree Teaching Material (printout whole set)
    • Dr. Eick's Short Introduction Decision Trees and Transparencies explaining the Information Gain Heuristic. and Dr. Eick's evaluation of decision tree techniques.
  • Neural Networks
    • Russel's Introduction to Neural Networks
    • Dr. Eick's additional transparencies discussing learning terminology and neural networks
    • Java Neural Network Animation and Neural Network Consulter (new!! --- I believe these animations are quite helpful for obtaining a general understanding on how neural networks work).
  • Introduction to Reinforcement Learning (many authors) & Learning for the XYZ World (discussion of Problem12 of Assignment3)
  • Han/Kamber's Introduction to Knowledge Discovery and Data Mining
• 2004 Bayesian Reasoning and Belief Network Transparencies
  • Review Probability Theory
  • Dr. Eick's Transparencies on "Naive Bayesian Classifiers" (only transparencies 1, 2, 13, 14 will be used in the lectures)
  • Dr. Eick's Computations in and with Belief Networks Transparencies
  • Lesser's Introduction to Belief Networks (explains d-separation quite well).
  • Slides 1997 AAAI Tutorial by Jack Breese and Daphne Koller on Belief Networks and Decision-Theoretic Reasoning (background knowledge, will not be used in the lecture)
• Soft Computing: Zadeh's Definition of Soft Computing & Introduction to Soft Computing (just print out the first 3 pages of the pdf-file that will be used in the lecture) --- Fuzzy Logic Software Tool (one of your class mates was instrumental in developing that tool).
• Foundations of AI (quite short; to be discussed in the class on Dec. 2, 2004)
• Yves Lesperance's Brief Introduction to Intelligent Agents, Dr. Eick's Shared Ontologies presentation.
• Russel Planning Transparencies (possibly will be used for a short "bird's perspective" introduction to planning)
• 2004: Topics covered and not covered in COSC 6368.
• Dec. 7, 2004 Review for the final exam; Solution Problem 16.

Textbook Coverage

Grading

Translation number to letter grades:
A:100-90 A-:90-86 B+:86-82 B:82-77 B-:77-74 C+:74-70
C: 70-66 C-:66-62 D+:62-58 D:58-54 D-:54-50 F: 50-0

Only machine written solutions to homeworks and assignments are accepted (the only exception to this point are figures and complex formulas) in the assignments. Be aware of the fact that our only source of information is what you have turned in. If we are not capable to understand your solution, you will receive a low score. Moreover, students should not throw away returned assignments or tests.

Students may discuss course material and homeworks, but must take special care to discern the difference between collaborating in order to increase understanding of course materials and collaborating on the homework / course project itself. We encourage students to help each other understand course material to clarify the meaning of homework problems or to discuss problem-solving strategies, but it is not permissible for one student to help or be helped by another student in working through homework problems and in the course project. If, in discussing course materials and problems, students believe that their like-mindedness from such discussions could be construed as collaboration on their assignments, students must cite each other, briefly explaining the extent of their collaboration. Any assistance that is not given proper citation may be considered a violation of the Honor Code, and might result in obtaining a grade of F in the course, and in further prosecution.

Course Exams

Midterm

Final Exam

2004 AI Qualifying Exam

• Heuristic Search (focusing on chapters 3-4 of the textbook and on material covered in the lectures); however, games will not be covered in Part2.
• Decision trees (see requirements 2004 midterm exam)
• Logical Reasoning (see requirements midterm exam)
• There will be questions concerning the contents of the following paper: Thomas G. Ditterich: "Machine Learning Research --- 4 Current Directions" (questions will not refer to material covered in sections 5.3 and 5.4 of the paper). 40-50% of the available points for Part2 of the QE will be allocated to this part.

Communication with the teaching staff

Please do not e-mail us with grading questions. If you want us/me to explain why I took points off, you can talk to me/us during office hours and directly after class.

Material COSC 6368 Fall 1999

Assignment Problems Fall 2002

Projects and Assignments Fall 2001

And finally: Frogland --- all about frogs