Abstract

Decision aids provided to foresters by forest management research are generally based on deterministic models. However, different sources of randomness may affect significantly the decisions to be taken. In this preliminary study, we consider the effect of uncertainty on the tree mortality factor for the even-aged stand management problem. This problem is modeled as a Markov Decision Problem (MDP), where solutions are defined as policies which specify the actions to perform for each state of the system. The algorithm Q-learning we apply in this paper, which belong to the field of reinforcement learning, is designed to automatically generate the optimal policy of an MDP on the basis of simulations. Preliminary experimental results are presented, that do not establish clearly the influence of stochastic factors on optimal solutions

Published in

Publisher: Silsoe Research Institute, Siloe, England

Conference

EWDA-04 European Workshop for Decision Problems in Agriculture and Natural Resources

SLU Authors

• Eriksson, Ola

  • Department of Forest Resource Management, Swedish University of Agricultural Sciences
    

• Wikström, Peder

  • Department of Forest Resource Management, Swedish University of Agricultural Sciences
    

UKÄ Subject classification

Forest Science


Permanent link to this page (URI)

https://res.slu.se/id/publ/5206