Open oyamad opened 6 years ago
So far the action x is assumed to be a 1dim continuous variable.
x
Implementing s_wise_max for discrete action will be straightforward. The issue will be how to design the ContinuousDP type.
s_wise_max
ContinuousDP
I think this would be really great. Being able to express mixed discrete/continuous problems and solve them efficiently would make this package much more valuable to my own research
So far the action
x
is assumed to be a 1dim continuous variable.Implementing
s_wise_max
for discrete action will be straightforward. The issue will be how to design theContinuousDP
type.