Fan Chung Graham
Algorithm Design and Analysis
TA: Alex Tsiatas
TA: Wenbo Zhao
Time & Place: Lectures M W 5-6:20pm Peterson Hall 102
Discussion Section Th 4-4:50pm Warren Lecture Hall 2111
Office Hours: Fan Chung Graham, APM 7101 W 2:30-3:30 pm.
Alex Tsiatas, EBU3B (CSE Building) B250A, T 12:00-1:30pm.
Wenbo Zhao, EBU3B (CSE Building) 4232, T 5:00-6:00pm.
This couses covers two main themes --
basic algorithms and
some recent developments on Internet algorithms. Also see the
Departmental CSE202 page.
The text book is
Algorithm Design by J. Kleinberg and E. Tardos.
Grading: 5 homework sets (20%), 1 midterm (30%) and 1 final (50%)
Weeks 1-2, Chapter 1, Introductory problems (Reading: Chapters 2-3. Also
check the reading list)
Chapter 4, Greedy Algorithms (covering 4.4-4.6)
Weeks 4-5, Chapter 5, Divide and conquer (covering 5.1-5.4)
Weeks 5-6, Chapter 6, Dynamic Programing (covering 6.1-6.4 and 6.6-6.8)
Note the coverage has been changed.
Midterm May 5, Wednesday
Weeks 7-8, Chapter 7, Network Flow (covering 7.1-7.3, 7.5-7.10)
Week 9, Chapter 11, Approximation Algorithms (covering 11.1-11.3)
(Reading: Chapter 8, NP and Computational Intractability )
Week 10, Local Search with additional material on PageRank algorithms
- Final Exam June 11, Friday, 7-10pm Peterson 102
Homework: All homework assignments should be handed in class
(before the lecture starts) at the
specified due dates:
Homework #1 (Wed. April 7)
Homework #2 (Wed. April 21)
Homework #3 (Mon. May 3)
Homework #4 (Wed. May 19)
Homework #5 (Wed. June 2)
The midterm and final will include problems very similar to those in homework assignments.
late homework will be accepted.
Due to the heavy load for our TA, not all of the homework problems will be
At least one problem from each set will be randomly chosen for grading.
Only the best four scores out of five homework assigments will be taken into account.
Note that the exam scores depend on the efficiency of your algorithm. For example, if the
algorithm has running time O(log n) but your algorithm is O(n2),
only get a very partial score.