Learning hierarchical observable Markov decision process models for robot navigation

2002 ◽  
Author(s):  
G. Theocharous ◽  
K. Rohanimanesh ◽  
S. Maharlevan
Keyword(s):  
Markov Decision Process ◽  
Decision Process ◽  
Robot Navigation ◽  
Process Models ◽  
Markov Decision

scholarly journals Risk aversion and risk seeking in multicriteria forest management: a Markov decision process approach

2017 ◽  
Vol 47 (6) ◽  
pp. 800-807 ◽  
Author(s):  
Joseph Buongiorno ◽  
Mo Zhou ◽  
Craig Johnston
Keyword(s):  
Risk Aversion ◽  
Markov Decision Process ◽  
Decision Process ◽  
Weighted Average ◽  
Basal Area ◽  
Risk Attitude ◽  
Expected Value ◽  
Process Models ◽  
Risk Seeking ◽  
Markov Decision

Markov decision process models were extended to reflect some consequences of the risk attitude of forestry decision makers. One approach consisted of maximizing the expected value of a criterion subject to an upper bound on the variance or, symmetrically, minimizing the variance subject to a lower bound on the expected value. The other method used the certainty equivalent criterion, a weighted average of the expected value and variance. The two approaches were applied to data for mixed softwood–hardwood forests in the southern United States with multiple financial and ecological criteria. Compared with risk neutrality or risk seeking, financial risk aversion reduced expected annual financial returns and production and led to shorter cutting cycles that lowered the expected diversity of tree species and size, stand basal area, stored CO2e, and old-growth area.

Sign in / Sign up

Export Citation Format

Share Document