compute_belief: Compute current belief state

In conservation-decisions/smsPOMDP: POMDP problem solver: When to stop managing or surveying cryptic threatened species ?

Description

Compute current belief state given past actions, observations and initial belief state

Usage

compute_belief(p0, pm, d0, dm, ds, V, Cm, Cs, state_prior, act, obs, disc = 0.95)

Arguments

p0	Local probability of persitance : P(extant/extant, survey or stop).
pm	Local probability of persitance if manage : P(extant/extant, manage).
d0	Local probability of detection : P(present/extant, stop).
dm	Local probability of detection : P(present/extant, manage).
ds	Local probability of detection if survey : P(present/extant, survey).
V	Estimated economic value of the species ($/yr).
Cm	Estimated cost of managing ($/yr).
Cs	Estimated cost of survey ($/yr).
state_prior	Initial belief state, vector of 2 values (belief state extant and extinct), between 0 and 1.
act	list of actions, values in c('Manage', 'Survey', 'Stop')
obs	list of observations, values in c('Seen', 'Not_seen')
disc	Discount factor used to compute the policy (default 0.95)

Value

Current belief state, vector of 2 values (belief state extant and extinct), between 0 and 1.

Author(s)

Luz Pascal

Examples

 ## Not run: 
#values for Sumatran tigers
pen <- 0.1
p0 <- 1-pen
pem <- 0.05816
pm <- 1 - pem
V <- 175.133
Cm <- 18.784
Cs <- 10.840
d0 <- 0.01
dm <- 0.01
ds <- 0.78193

#Initial belief state
state_prior <- c(0.9,0.1) #extant : 0.9, extinct : 0.1

#previous actions and observations
ac <- c('Manage','Survey','Stop')
ob <- c('Not_seen','Not_seen', 'Seen')

compute_belief(p0, pm, d0, dm, ds, V, Cm, Cs, state_prior, ac, ob)
 
## End(Not run)

conservation-decisions/smsPOMDP documentation built on Oct. 27, 2020, 10:44 p.m.