MDPs and POMDPs in Julia - An interface for defining, solving, and simulating fully and partially observable Markov decision processes on discrete and continuous spaces.
OTHER License
Bot releases are hidden (Show)
Published by zsunberg about 2 years ago
Published by zsunberg over 2 years ago
Added documentation about transition
, gen
, and @gen
Published by zsunberg almost 3 years ago
Updated documentation with a QuickPOMDPs- and distributions-centric guide.
Published by zsunberg about 6 years ago
Published by zsunberg about 6 years ago
Published by rejuvyesh about 6 years ago
Published by MaximeBouton about 6 years ago
Published by zsunberg about 6 years ago
Compatible with Julia Version 1.0
Squashed some names together, like initial_state -> initialstate, action_index -> actionindex, etc.
Published by MaximeBouton about 6 years ago
Last version compatible with julia 0.6
Published by zsunberg over 6 years ago
Added a couple new solvers, removed default discount
implementation.
Published by zsunberg over 6 years ago
Added AEMS and FIB Solvers
Published by zsunberg almost 7 years ago
Much better feedback is given to the programmer when generate function synthesis fails.
Published by zsunberg almost 7 years ago
Published by zsunberg almost 7 years ago
Sometimes very specific method signatures in POMDPs.jl make it more difficult to define methods for a problem definition, e.g. if isterminal(p::POMDP{S,A,O}, s::S)
is defined in POMDPs.jl, then isterminal(p::MyPOMDP, s)
will be ambiguous. This release removes overly specific definitions to make things easier.
Published by zsunberg about 7 years ago
Published by zsunberg about 7 years ago
BasicPOMCP replaces POMCP in the list of solvers.
Also made the default implementations of the generate functions more general to try to prevent ambiguity errors.
Published by MaximeBouton over 7 years ago
Published by MaximeBouton over 7 years ago
Published by zsunberg over 7 years ago
fix #140 (adding sampletype()), many other small fixes