Entropy-regularized Point-based Value Iteration

Entropy regularized point-based value iteration solver for the POMDPs.jl framework.

Installation

From the julia REPL,

]add [email protected]:hdelecki/EntropyRegularizedPBVI.jl.git

Usage

using EntropyRegularizedPBVI
using POMDPModels

pomdp = TigerPOMDP()

solver = ERPBVISolver(max_iterations=10, ϵ=0.1, λ=0.1, verbose=true)

policy = solve(solver, pomdp)