作者: Byron K. Williams , James D. Nichols
DOI: 10.1007/978-1-4899-8041-0_4
关键词:
摘要: The previous three chapters of this book have been devoted to specific components informed decision processes: objectives, potential actions, model(s) predicting system change and response monitoring provide estimates status. final component an process is a solution algorithm, providing means for deciding which action take. Optimization algorithms objective transparent approach select the that will do best job meeting objectives. Static optimization provides problems are not iterative, we examples one or more variables (variables actions). Many in natural resource management viewed as dynamic, they iterative require decisions repeated through time. In dynamic problems, made at point time expected influence state next step, thus influencing state-dependent For any decision, must consider all subsequent steps horizon problem. addition being most characterized by substantial uncertainty, extended deal with several sources uncertainty. An important source uncertainty about how responds may develop multiple models characterize Adaptive solutions only but anticipated reduction future decisions. output algorithm frequently graph table recommended actions values variables. Decision thresholds defined simply locations space where small value variable produces optimal action.