CSE 8843

COMPLEXITY OF SEQUENTIAL AND PARALLEL ALGORITHMS

Prof. Ioana Banicescu.

HOURS:

CLASS: Mon/Wed 3:30 - 4:45, Butler 103.

OFFICE: 320 Butler Hall, Tue/Thu 2:00 - 3:00; other times by appointment only. Email: ioana@cse.msstate.edu.

325-2756

TEXT BOOK:

• Algorithms: Sequential, Parallel and Distributed, Kenneth Berman and Jerome Paul, Thomson - Course Technology, 2005.
• Introduction to Algorithms, second edition. Thomas H. Cormen, Charles E. Leiserson, Ronald L. Rivest and Clifford Stein. The MIT Press, 2001.

RECOMMENDED SELECTED READINGS FROM:

• John Savage, Models of Computation - Exploiting the Power of Computing, Addison-Wesley, 1998. (Supplementary Readings).
• Joseph JaJa, An Introduction to Parallel Algorithms, Addison-Wesley, 1992.
• Russ Miller and Laurence Boxer, Algorithms Sequential and Parallel-A Unified Approach, Prentice Hall Inc, 2000.
• John Reif, Synthesis of Parallel Algorithms, Morgan Kaufmann 1993.

OTHER REFERENCES:

• Thomson Leighton, Introduction to Parallel Algorithms and Architectures, Morgan Kaufmann 1992.
• Vipin Kumar, Ananth Grama, Anshul Gupta, George Karypis, Introduction to Parallel Computing, Design and Analysis of Algorithms The Benjamin/Cummings Publishing Co., Inc. 1994.

PREREQUISITES:

The required prerequisites are listed in the official university bulletin. Graduate level courses in algorithms, parallel algorithms and high performance computing are desirable.

OBJECTIVES:

The goal of this course is to familiarize the students with the general methodology and specific techniques used in the analysis of sequential and parallel algorithms. The course will be focused on theoretical aspects of algorithms' complexity as well as on the analysis and evaluation of their effectiveness in parallel applications' performance.

COURSE REQUIREMENTS:

This course examines a variety of theoretical issues that have important practical consequences in the design of sequential and parallel algorithms. We emphasize the importance of asymptotic analysis and the common methodology used in the analysis of sequential and parallel algorithms. We will analyze the factors that influence the performance of algorithms on parallel machines and their implications on scientific applications of interest. We will survey major design strategies used in the complexity analysis of different algorithms. New models for parallel computation will be studied and critically evaluated. The complexity and limitations of these models will also be discussed.

GRADING:

There will be several assignments, test exams and a comprehensive final exam. The final grade will be based on all of the above (assignments and test exams: 60%; comprehensive exam: 40%).

ANNOUNCEMENTS:

Students are responsible for checking the announcements regularly click here.

COURSE OUTLINE:

The course will cover material from the following topics:

Fundamentals: asymptotic analysis,design and analysis of sequential algorithms(average case and amortized analysis, solving recurrences).

Major Design Strategies: divide-and-conquer, greedy method, dynamic programming, backtracking and branch-and-bound.

Analysis of a few sequential and parallel algorithms of interest (sorting, graph, matrix, search algorithms for discrete optimization, dynamic programming, and others).

Computational complexity (space-bounded class, time-bounded class, P-complete problems, NP-complete problems, NP-hard problems, nondeterministic algorithms, etc), and circuit complexity.

Probability and average complexity of algorithms; Heuristic and approximate algorithms.

Introduction to Lower Bound Theory.

Performance analysis and evaluation of parallel systems - scalability analysis (performance metrics, execution time, speedup, efficiency, cost, scalability, isoefficiency and cost effectiveness).

Models of parallel computation (PRAM, LogP, BSP, etc.) and their limitations.

Space-Time tradeoffs and Memory-Hierarchy tradeoffs.

ACADEMIC HONESTY, ETHICS IN RESEARCH AND SCHOLARLY ACTIVITIES:

Students are expected to follow MSU as well as the CS department's policies on academic misconduct and on academic honesty with regard to homeworks, projects, exams, papers. These policies can be found on the MSU ethics in research and other scholarly activities and Computer Science and Engineering Departmental policy regarding professional conduct pages respectively. Also, as of fall semester 2007, Mississippi State University has an approved Honor Code that applies to all students, and reads as follows: As a Mississippi State University student I will conduct myself with honor and integrity at all times. "I will not lie, cheat, or steal, nor will I accept the actions of those who do". Upon accepting admission to Mississippi State University, a student immediately assumes a commitment to uphold the Honor Code, and accepts responsibility for learning and following the philosophy and rules of the Honor Code. For additional information, please see the Honor Code.

---

ATTENDANCE AND ADD/DROP POLICIES:

Attendance is required. Roll will be taken, and absences will be listed on midterm and final grade reports. Attendance may influence grades at border line. Makeup exams are allowed only in special circumstances where absences are certified by the Dean of Students office. At the beginning of the semester, students are encouraged to also review the Add/Drop policy.

Students who are auditing this course must attend at least 75% of the class lectures to avoid a grade of F.