SIMBREED: Simulated Animal Breeding as a Computational Metaphor for Optimizing Agricultural Systems
|
 |
|
Post a Comment
|
|
|
|
|
CONTRIBUTORS:
|
|
|
JOURNAL:
|
|
|
YEAR:
|
2004
|
|
PUB TYPE:
|
Journal Article
|
|
SUBJECT(S):
|
Breeding; combinatorial computation; heuristics; SIMBREED; optimization
|
|
DISCIPLINE:
|
Animal Sciences
|
|
HTTP:
|
http://www.ics.uplb.edu.ph/node/108
|
|
LANGUAGE:
|
English
|
|
PUB ID:
|
103-444-110
(Last edited on
2008/07/19 01:48:19 GMT-6)
|
|
SPONSOR(S):
|
|
|
ABSTRACT:
We present a breeding-inspired computational method for solving optimization problems in agricultural systems. We discuss SIMBREED, a set of computer heuristics we developed, that simulates and uses animal breeding as a computational metaphor for finding solutions to combinatorial optimization problems such as the general diet formulation problem (DFP). We use DFP as a representative problem to demonstrate that SIMBREED algorithms can efficiently solve complex optimization problems. We present solutions to DFP by mapping morphogenetic traits to DFP variables and satisfy the objective function via natural selection or selective breeding, with small chance of mutation. We compare the DFP solution found by ADBASE, a deterministic algorithm, to those found by SIMBREED algorithms. We also compare the performance of two SIMBREED algorithms: simulated natural selection (EVOLVE) and simulated selective line breeding (BREED). Results show that SIMBREED can be used to find solutions to complex combinatorial optimization problems such as the DFP better than a deterministic algorithm and that the simulated selective line breeding performs better than the simulated natural selection.
|
|
|
|
STATISTICS
|
|
Click on # to view
|
|
Citations
|
|
0
|
|
References
|
|
0
|
|
Comments
|
|
0
|
|
Quality
|
|
0/0.00
|
|
Interest
|
|
0/0.00
|
|
View(er)s
|
|
2/93
|
|
|
|
|
|
|
| Prev |
Next |
|