Computer Science 142 and Mathematics 167
Computational Complexity Theory (a.k.a. "Theocomp")
Syllabus, Fall 2008
Professor: Ran ("RON") Libeskind-Hadas
Office: Olin 1245
Phone: x18976
E-mail: hadas@cs.hmc.edu
Office Hours: Please see my schedule
Course Time and Location: Mondays and Wednesdays, 4:15-5:30 PM, in
Jacobs B134
Course Homepage:
www.cs.hmc.edu/courses/2008/fall/cs142
What Is This Course About?
This course is primarily about computational complexity theory, the
study of the "hardness" of computational problems. However, we begin
with a brief section on computability theory (some review from CS 81
but mostly new material), since a firm grounding in this area is
necessary to understand some of the issues in complexity theory.
The course is
divided into two parts: Part 1 of the course will be lecture-based
and will cover the foundations of computability and complexity.
In Part 2 of the course,
you will select a topic in complexity theory, give a presentation to
the rest of the class on your topic, and assign and grade homework
problems on your presentation.
Here is a tentative list of topics:
Computability Theory
- Turing Machines and decision, function, and enumeration
problems
- Godelization and equivalence with restricted models of computation
- Diagonalization, the halting problem, reductions, and Rice's
Theorem
- Undecidability of Tiling Problems
- State minimization
- The Recursion Theorem and its implications
Computational Complexity Theory
- Time and space measures on Turing Machines and other models of computation
- Polynomial-time and log-space reductions, P, NP, and NP-completeness
- Cook's Theorem
- Strong NP-completeness
- NP and co-NP
- The complexity of approximation
- Function problems and FNP
- The polynomial hierarchy and PSPACE-completeness
- Savitch's Theorem
- P-completeness
- Logspace and NL-completeness
- Randomized algorithms and complexity classes
- The classes IP and ZKP
- The speedup, hierarchy, and gap theorems
Part 2 of the course will comprise student presentations of material
related to this course. The objective of this part of the course is to
develop the ability to read original research papers and present the material
clearly in class.
You will work with Ran in advance to select the topic, find the appropriate
papers to read, and develop your in-class presentation.
In addition, you will be asked to assign approximately two homework problems on
your topic and you will grade these problems. These homeworks scores
will count as required regular homework for your classmates.
Texts
There is no textbook for the course. Lectures will be self-contained.
Some books will be placed on reserve in the library and this will be
announced in class.
Attendance and Participation
On-time attendance in every class is absolutely required to pass the course.
If you are sick or have a
special reason for missing class, please contact Ran before the class
that you will miss. Otherwise, you are expected to be in class.
Please make sure to arrive on-time as a courtesy to the instructor and your
classmates. Participation in class is also expected and is a component of the
course grade.
Assignments
There will typically be one assignment each week, given on Wednesday and
due the following Wednesday.
All problem set submissions must be typeset, preferably using LaTeX.
Assignments will be posted in both pdf and LaTeX source formats on the
course web page.
Exams
There will be one exam in this course, an open-notes
take-home exam on the material in Part 1 of the course.
There will be no final exam.
Grades
The components of the course are worth the following:
Homework in Parts 1 and 2 of the course: 60%
Take-home Exam: 20%
Part 2 presentation: 20%
Note that attendance is not factored into this breakdown, but is
required in order to pass the course.
Collaboration Policy
Collaboration on homeworks is encouraged. This means that you may
discuss approaches to solving a problem with anyone in the class.
On each submitted homework, please indicate the names of all people
with whom you discussed your solutions. Unless indicated otherwise,
no written sources (other than your own notes) should be consulted in
solving homework problems.
There should be no collaboration on the take-home exam.
All parties, including off-campus students,
are expected to follow the Harvey Mudd Honor code.