Skip to main content

Table 1 Parameter settings and brief descriptions

From: Learning versus optimal intervention in random Boolean networks

ParameterValueBrief Description
R1400Rule population size
γ0.71Discount rate
θmnaN+1Min. number of actions in match set
P#0.33Probability of hash
pI0.0Initial payoff
εI0.0Initial error
FI0.0Initial fitness
ε010Error threshold
θga5.0Genetic algorithm frequency
θdel20.0Deletion threshold
β0.2Affects update of p, ε, and action set size for classifiers
α0.1Affects fitness updates
ν5.0Affects fitness updates
χ0.8Likelihood of GA crossover operation
μ0.05Likelihood of GA mutation operation
δ0.1Mods. effect of fitness on the ‘deletion vote’ of a classifier
θsub30.0Subsumption threshold
pexplr0.5Likelihood of exploring
doAsSubsumpt.truePerform subsumption in the action set?
doGaSubsumpt.truePerform subsumption in the GA?