This short series of lectures deals with algorithmic techniques based on two well-know paradigms for designing approximation algorithms: the Primal-Dual Schema (PDS) and the Local-Ratio Technique (LRT).

The first lecture is an introduction to PDS and LRT. Each subsequent lecture covers a particular algorithmic technique related to the two paradigms. Even though each lecture will typically focus on one or two problem-specific results, we will always emphasize the underlying principles at work. It is worth noting that no previous knowledge on PDS or LRT will be assumed.

Below is a tentative schedule for the lectures. Any additional material and future updates will be posted in this space.

Date Topic Relevant papers Slides
Nov 29 Intro to PDS and LRT and sketch of equivalence [V Ch. 13] [BBFR] ppt
Dec 13 Randomized local ratio [B] [ACN]
Jan 17 Lagrangian relaxation [KPS] ppt
Feb 7 Dual fitting and factor revealing programs [V Ch. 12] [IMM]
Mar 6 On-line primal-dual algorithms [BJN]