COSC 6368 --- Artificial Intelligence Fall 2017
( Dr. Eick )
Purpose of this Website
This website intends to satisfy the information requirements of
two independent groups:
- Students that take the graduate AI class
- People that want to find out what AI is,
what its subfields are,
and how its technologies, techniques, and
methodlogies can be used in industrial, government, and
research-oriented environments.
If you have any comments
concerning this website, send e-mail
to: ceick@aol.com
Basic Course Information
2017 COSC 6368 Syllabus
class meets: TU/TH 2:30-4p
class room: SEC 202
Instructor: Dr.
Christoph F. Eick
office hours (573 PGH): TU 11:15-12:30p TH 4-4:45p
TA: Reza Fathi
TA office hour: TU noon-1p and TH 1-2p in 313 PGH
TA Email: taemail172@gmail.com
Reza's 6368 Webpage
office: PGH 313
cancelled classes: none at the moment
makeup classes: none at the moment
lectures taught by others:
Topics Covered in COSC 6368
The course will give an introduction to AI and it will cover Problem Solving (covering
chapter 3, 4 in part, 5, and 6 in part,
centering on uninformed and informed search , adversarial search and
games, A*, alpha-beta search, evolutionary computing, game theory (chapter 17 and other matrial), and
constraint satisfaction problems, discussion course Project 1)), Planning(covering chapters 10 and
11 in part), Learning (covering learning from examples (chapter 18), deep learning (extra material) and
reinforcement
learning (chapter 21, chapter17 in part;), Discussion Project2), Reasoning and Learning in Uncertain
Environments (covers chapters 13, 14, 15 in part, and 20 in part, centering on ?basics? in probabilistic reasoning,
na?ve Bayesian approaches, belief networks and hidden markov
models (HMM)) and might, if enough time might
briefly discuss Robotics, Philosophical Foundations of AI.
The course will cover Chapter 1, 2, 3, 4, 5, 6, 10, 11, 13, 14, 15, 17, 18, 20, 21, and might, if
enough time, partially cover
Chapters 25, and 26 of the Stuart Russel & Peter Norvig book.
Course Materials
Required Text:
- S. Russell and P. Norvig, Artificial Intelligence, A
Modern Approach, Third Edition,
- Prentice Hall/Allyn&Bacon, 2010,
-
Link to Textbook Homepage.
News COSC 6368 Fall 2017
- COSC 6368 letter grades and grade summaries should be available at Reza's website
Friday afternoon,
and very detailed grade reports will
be posted on Reza's Website by Saturday, we hope!
- The final exam will not be returned to students; however, you can see
you final exam Thursday, December 14, 3-4p or Thursday, January 18, 2-3p (some solution
sketeches for the final exam can be found near the bottom of this webpage).
- We were a little generous in computing the early submission bonus for Project1; we
are now adding 4 points to your score which means a bonus of 5% for students who have
the current average score of 80; we do the same for Project2. The raw scores average for both
projects are now between 79 and 80 for both projects; as mentioned earlier, these score will
still be curved when converting them into number grades.
Important Dates in the Fall 2017 Semester (updated!)
Fr., October 6/Mo., October 9: Submission deadline Project1
Tu., October 17: Submission deadline Homework1
Tu., October 24: Midterm exam
Su., November 19/We., November 22: Submission deadline Project2
Fr., December 1: Submission Deadline Homework2 (optional for students who gave a presentation)
Th., November 30: last lecture COSC 6368
Th., December 7, 2p: Final exam (Review List Final Exam)
All deadlines are 11p, unless explicitly specified, otherwise.
2017 Course Projects and Homeworks
Project1: Using AI to Solve Travelling Salesman Problems
(individual project)
Homework1 (group homework;
H1 Groups)
Project2: Learning Paths with Q-Learning and SARSA (PD-World, Q-Learning and SARSA
Example for a Pickup World)
Homework2 (Solution
Sketches Homework2 for Problems 1, 3, 4)
Academic Honesty: I also like to emphasize that Project1 and Project2 are
individual projects. Collaborating with other
students in COSC 6368 and using code your class mates wrote is strictly disallowed. Reporting
false results
or solutions that were not obtained by running your own programs is also a serious academic honesty violation. Finally, not reporting external software you used in your implementation is
also an academic honesty violation.
2017 Reviews and Exams
October 17 Review for Midterm Exam
(Review List; Solution
Sketches Midterm 2017)
November 30: Review for the 2017 Final Exam (Solution Sketches Nov. 30, 2017 Review)
Final Weights of the Course Elements in 2017
2 Exams:54% (24%+30%), Homework1: 5%,
2 Projects: 35% (17.5%+17.5%), Presentation/Homework2: 4%, Attendance: 2%
COSC 6368 Lecture Transparencies
- 2017 Introduction to AI and Course Information
COSC 6368 (will be
used for first lecture) and Intelligent Agents (briefly covered in 2017).
- 2017 Search Transparencies:
- Search1 (Classification of Search Problems, Terminology, and Overview
),
Search2 (Problem Solving Agents),
Search3 (Heuristic Search and Exploration),
Search4 (Randomized Hill Climbing and Backtracking; not covered in textbook),
Kamil on Backtracking and Mazes,
Search5 (Games; Russel transparencies for Chapter 6; will not cover
transparencies that discuss card games),
Search5a (Brief Discussion of Bridge and Man vs. Machine Game Contests),
Search6 (Russel slides Constraint Satisifaction Problems (CSP); we will cover
the first 32 slides in part),
Search6a (Dhar & Quale paper on Dependency Directed Backtracking (DDBT),
a "better" form of backtracking for CSPs; the idea of DDBT will be discussed in the lecture,
but not the whole paper!),
Search7 (Discussion of Greedy Search and A*; not covered in the lecture, but in the review
in October).
- Suggestions for Solving the Rook+King vs. King Endgame (WRKBK) Problem.
- 2017 Teaching Material Evolutionary
Computing (EC): EC1: Introduction
to Evolutionary Computing and EC2:Example: Using EC to Solve Travelling
Salesman Problems, Eiben-Smith Introduction to EC (they call 'EC': 'EA')
- 2016 Game Theory Slides: G1: Introduction to Gametheory (USC Economics slide show
by Shivendra Awasthi (???), will be used in the lecture) and G2: Mo Tanweer Mohammed's Introduction to Game
Theory (not used in the COSC 6368 lecture).
- 2017 AI Planning Slides: PL1: Sketch How a STRIPS-like Planning System Solves a Block's World Problem (was done
on the white board in the lecture on Tu., Oct. 6 (Sheet1,
Sheet2)),
PL2 (Blythe, Ambile, Gil (USC): Introduction to Planning slides, covered in the lecture),
PL3 (Jussi Rintanen's ECAI 2014 Planning Tutorial; we will discuss
slides 1-7, 10-14, 31, 51-52)
- 2017 Machine Learning Transparencies:
- Quick Introduction to
Machine Learning.
- Reinforcement Learning: RL1 (Introduction to Reinforcment Learning),
RL2 (Using Reinforcement
Learning for Robot Soccer), RL3 (Kaelbling's RL Survey Article: read
sections 1, 2, 3, 4.1, 4.2, 8.1 and 9 discussed in the lecture)
- Decision Trees: DT1 (Dr. Eick's Introduction to Decision Trees,
DT2 (Russel
Decision Tree Slides; only the first 6 transparencies
will be used)
- Neural Networks: NN1 (Russel's Introduction to Neural Networks),
NN2 (Dr. Eick's additional NN slides),
NN3 (Java
Neural Network Animation) and NN4 (Neural Network Consulter).
- A Short Introduction to Deep Learning (by Fabio Gonzalez, National University of Colombia)
- 2017 Decision Making and Reasoning in Uncertain Environment
Transparencies
- 2017 Last Words
- 2009 Logical Reasoning Transparencies:
- 2006/2009 Soft Computing Transparencies
- A quick look to Knowlege-based Systems
- Foundations of AI (quite short; to
be discussed in the last class of the semester)
- 2009: Topics covered and
not covered in COSC 6368.
- Dec. 7, 2004 Review for the final exam;
October 19 "Lecture"
Please view the following 3 videos which "form" the October 19 lecture:
- Siraj Raval: How to use Q Learning in Video Games Easily (7 minutes)
- Richard Sutton: Deconstructing Reinforcement Learning (about 50 minutes)
- Eric Guimarães:Demo Q-Learning in a GridWorld(2 minutes)
Results of the Nov. 28 Questionnaire
Maybe, due to Dr. Vilalta's final exam on Nov. 29, and only 17 students filled
out the questionnaire; here are the findings:
- 7 students liked Project1, 4 were neutral, and 2 disliked the project; 9
students liked Project2, 3 were neutral, 1 disliked the project; the remaining 3
students did not express a clear opinion on this matter. 5 students expressed
the opinion that Project2 was more demanding than Project1: 2 thought
that Project2 was too demanding; 2 thought that having a
challenging project sometimes help to uncover hidden talents and helps
you to grow beyond your own boundaries;
the other 12
students did not voice an opinion if Project2 was more demanding than Project1.
3 thought that Project1
was not challenging enough, but the other 14 students' assessment do not make
such a statement.
- Compared to other courses, 5 students found the projects in the
course more demanding,
5 students found them equally demanding, and 1 found them less demanding; 5
liked the COSC 6368 projects more than the projects in other courses, and 2 were neutral.
- 14 found the peer reviews helpful, 2 were neutral, and 1 thought that they
are not helpful.
- 13 found group presentations to be a positive course element, 2 were
neutral,
and 2 thought that they should be replaced by videos and/or homeworks.
2 student would have prefered to assign predefined topics to each group,
and to impose constraints on what is covered in the group presentation.
- One student said that COSC 6368 lectures are sometimes difficult to understand,
and suggested to identify a "prerequisite for each lecture" that
allows students to "better prepare for lectures".
- 3 students would prefer to get a clearer understaning what is evaluated
in the demos before hand, 3 student thought that demos should
focus more on AI aspects
of the project (and not so much on software design, user interfaces,
visualization, efficiency of code and algorithms, etc.); one student found it postive that we, in contrast to some
other courses, evaluate if the project software solves the problem at hand, which he/shefeels leads to a "fairer" evaluation; two other students who visited me during my office hours voiced
the same opinion.
2 students did not like of being asked to delete
some code and being asked to rewrite it (the goal of that was to assess
if a student actually had written the particular software, and a few students
were randomly chosen for this task, according to Reza).
Prerequisites
Students are expected to have the following background:
- The COSC 4368 prerequisite will not be enforced; you can take COSC 6368 without having taken COSC 4368 or
similar courses! COSC 6368 is a self-contained course: no prior knowledge of AI is required!
- 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.
- Familiarity
with the basic concepts of discrete probability theory. However, it suffices to read and understand
Chapter 13 of Russell and Norvig which will also.
- Students should have basic
programmig skills. Some assignments
require programming; however, students can choose the programming language they like the most.
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.
2017 COSC 6368 Student Presentations
Each student will either participate in a presentation or submit Homework2; we allocated
only approx. 3 minutes for each student. Presentations will be given by groups of (2)3-4 students. Presentations
will discuss general AI topics; others will conduct a search
for things that are useful for the AI course and summarize their findings; others will demo AI
tools, AI applications or report about AI contests.
Advice on the presentations themselves: Practice your presentation to make sure
that you stay within the time-limit associated with your presentation. Each group member should present a part of the
presentation! Finally, groups should introduce their
members at the beginning of the presnentation and the first slide of the presentation
should contain the names and
a photo of the presenters, the presentation title, and whatever else you like to put on the first slide!
2017 COSC 6368 Presentation Schedule
September 21, 3:34p: #1: Suchismitha Vedala, Roopa Reddy Rajala and Charan Teja
Sakhamuri: About the International Automated Negotiating Agent Competition (11 minutes).
September 28, 3:25p: #2: Kinjal Kotadia, Manasvi Thakkar, Maksim Egorov and Ayzha Ward:
Presentation and Evaluation of the AAAI
2016 Best Paper Award Paper Bidirectional Search That is
Guaranteed to Meet in the Middle (13 minutes, followed by a 7 minute discussion
of the paper)
October 3, 3:30p: #3: Michael Bremner, Feng Guo, Xin Zhou and Ai Zhuo:
Angry Bird AI Competition (13 minutes)
October 31, 3:33p: #4: Shah,Ashna Milind, Hoang,Huy Thai and Banerjee, Romita: AI Planning
Tools (11 minutes)
November 16, 3:20p: #5: Lavanya,s.s; Yashwant, Jyothi and Kiranvarma: Outwitting
Poachers with AI (13 minutes; https://www.engadget.com/2017/05/21/drones-ai-help-stop-poaching-africa/ and https://www.nsf.gov/news/news_summ.jsp?cntn_id=138271)
November 16, 3:33p: #6: Arjun Subramanyam Varalakshmi, Qian Qiu and Chonghua Li:
Mastering the game of Go without Human Knowledge (11 minutes)
November 28, 3:20p:
#7: Akhil Talari, Mohit Kaduskar, Navya Doddapaneni:
"Artificial Intelligence for Speech Recognition" (11 minutes).
#8: November 28: 3:32p: Konreddy,Deepika Reddy, Priscilla Roy Imandi, Eric Jiang and Abheesta Reddy: "What does Corporate America think about AI" (13 minutes).
#9: November 30, 2:40p: Goutam Venkatesh, Siva Uday Sampreeth Chebolu, and Dinesh Reddy Bethi:
"On Humonoid Robots" (11 minutes).
#10: November 30, 2:52p: Anusha Nemilidinne, Harshitha Thallaparthi, and Chethana Dukkipati:
"Google Street View" (11 minutes).
Grading
The course will have a midterm and a final exam, Students will be responsible for material covered in the
lectures and assigned in the readings. All homeworks and
project reports are due at the date specified. No late submissions
will be accepted after
the due date. This policy will be strictly enforced.
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 project reports
are accepted. 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
approach or 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 grade reduction, obtaining a grade of F
in the course, and in further prosecution.
2016 Course Projects and Homeworks
Specification Course Project1 (Individual Project; last updated
on September 26; due on Th., October 6, 11p;
Project1
Training Benchmark)
Homework1 (individual homework; due on Mo., October 17, 11p)
Project2: Learning Paths from Feedback Using Q-Learning
(Group Project, PD-World; 2016
Project2 Groups; due on Nov. 19, 11p)
Homework2 (individual homework, due Nov. 30, 11p; except problem 6)
2016 Course Exams and Reviews
Review1 October 18, 2016
Review List Midterm Exam October 25, 2016
Review2 on December 1, 2016 (More solutions)
Review List Final Exam December 8, 2016, 2p
Previous Course Exams
Midterm
Review Sheet for 2001 Midterm Exam
2004 Midterm Exam with some Solutions
2016 Midterm Exam with Solution Sketches
Final Exam
COSC 6368 2017 Final Exam Fall with
Solution Sketches (in Word)
COSC 6368 Final Exam Fall 1999 (in Word)
COSC 6368 Final Exam Fall 2001 (in Word)
Fall 1999 AI PhD Qualifying Exam (in html)
last updated: December 14, 2017, 11p.
And finally: Frogland --- all about frogs