Optimization of Large Join Queries

Optimizing queries is one of the hardest problems in database field since exhaustive searching of an optimal query execution plan in a large search space is impractical. Genetic Programming is an evolutionary algorithm that makes use of stochastic decision making method for optimization problems. In pure Genetic Programming approach, initial population is generated randomly without any considerations on the generated individuals. We propose the use of heuristics in generating initial population for the genetic runs so that we are starting the process with relatively good individuals, expecting to obtain an optimal query execution plan earlier.
Ċ
Paper.pdf
(230k)
Ko Ko Naing,
Feb 10, 2009, 9:20 PM
ċ
Presentation.zip
(1565k)
Ko Ko Naing,
Feb 10, 2009, 9:21 PM
Ċ
Thesis.pdf
(1551k)
Ko Ko Naing,
Feb 10, 2009, 9:19 PM