September 10

Algorithm Design

Design
  Algorithm*
  Software*
  Game*
  Database
  Network

Algorithm design is influenced by complexity analysis
of the underlying problem.

Chapter 1:
5 problems:
Polynomial:
Interval Scheduling: Greedy Algorithm
Weighted Interval Scheduling: Dynamic Programming
Bipartite Matching: Network flow
NP-complete:
Independent Set
PSPACE complete:
Competitive Facility Location

Defining descriptions of instances of those problems
Introduction to DemeterF

The Specker Derivative Game: SDG

Playing games with algorithmic problems.
CNF
types (length, positive)
derivative
agents

==========================
September 17:

Graph Algorithms
http://www.itasoftware.com/careers/puzzles07.html#SligBladeRunner

Topological sorting

Fast distributed algorithms: MapReduce

DemeterF

Hw 2:

=====================================
September 23:

Greedy algorithms
  Dijkstra Shortest Path 
Merge Sort
Analysis of algorithms

Good architectures for algorithms:
Architectures for DemFGen and DemeterF
   Open Class Architecture (using nonempty behavior file)
   Implicit Invocation Architecture (using empty behavior file;
     use traversals to inject functionality)
The important classes of the DemeterF library
illustarted with BST.

   
Code organization: using a library

slides:
03*.ppt : topological sorting (36 +)
04greed.ppt Dijkstra 35+
04demo*.ppt demo Dijkstra 
05divide*.ppt Mergesort 1+


Oct. 1
============================================

What we do in this course
An algorithm is made up of flows through an object.
hw 1 discussion
Approaching SDG with randomized algorithms
02analysis.ppt
Dijkstra 04demo-dijkstra.ppt
05divide-and-conquer.ppt
DemeterF

Oct. 8
=============================================

Where we are
Dynamic Programming (-14)
Review: Approaching SDG with Randomized Algorithms
Homework 4
13randomized (15-19, 21-28)

What we did not do last time
02analysis.ppt
Dijkstra 04demo-dijkstra.ppt
05divide-and-conquer.ppt
DemeterF

Nov. 5
===============================================

Midterm review
Project / SDG components
NP-completeness proof: revisit
Network Flow

Nov. 12
========================================
see plan-second-half
/home/lieber/.www/courses/csg113/f08/lectures

/home/lieber/.www/courses/csg113/f08/solutions/4/charu4/Part3B

http://www.ccs.neu.edu/research/demeter/DemeterF/java-src/DemeterF/edu/neu/ccs/demeterf/demfgen/lib/RBNode.java.html

/home/lieber/.www/courses/csg113/f08/preparation
-rw-r--r--   1 lieber   faculty      183 Nov 12 12:25 pseudo-polynomial
-rw-r--r--   1 lieber   faculty     1371 Nov 12 11:49 maximum-versus-maximal-better
-rw-r--r--   1 lieber   faculty       48 Nov  9 22:30 ideal-world
-rw-r--r--   1 lieber   faculty      630 Nov  9 21:49 dynprogr-inspired

Nov. 19
=====================================================

Discussion of document due on Dec. 1
Symmetrization technique
Analyzing the running time of DemeterF programs
Review of DemeterF programs (Bryan's slides)

Studying your algorithms: nov14

put the new ones here:
/scratch/lieber/csg113-f08-sdg/nov19

NP, NP-complete defs.
Reviewing intractability
Proof that most break-even prices separate P from NP.

Red-black trees; self balancing binary search trees

Dec. 3
=========================================================

Contention Resolution: An application of randomization
  (and SDG derivative analysis)
Another class of algorithms: Linear Programming
  (The game inside the game)
Symmetrization
Algorithm review
Reductions
Extra credit homeworks
Proof that most break-even prices separate P from NP.


Dec. 10
==========================================

Algorithms for playing with algorithmic problems
  Buying derivatives in the store

Review slides (in dec10 directory)
  Some new material, in the context of the old.

Determining the IQ of your components