This paper demonstrates how genetic algorithms can be used to discover the structure of a Bayesian network from a given database with cases. The results presented, were obtained by applying four different types of genetic algorithms — SSGA (Steady State Genetic Algorithm), GAeλ (Genetic Algorithm elistist of degree λ), hSSGA (hybrid Steady State Genetic Algorithm) and the hGAeλ (hybrid Genetic Algorithm elitist of degree λ) — to simulations of the ALARM Network. The behaviour of these algorithms is studied as their parameters are varied.
|Title of host publication||Learning from Data|
|Subtitle of host publication||Artificial Intelligence and Statistics V|
|Editors||D. Fisher, H. Lenz|
|Place of Publication||New York|
|Publication status||Published - 1996|