r^k, s^k = g(s^{k-1}, a^k) followed by p^k, v^k = f(s^k).
References
Schrittwieser et al. (2020), arXiv:1911.08265, Methods (Search) and Equation (1). Read from the ar5iv rendering.
Examples
dynamics <- function(s, a) list(0.5, s + a)
Mzrecur(state = 0, action = 1, dynamics = dynamics)
#> $state
#> [1] 1
#>
#> $reward
#> [1] 0.5
#>
#> $policy
#> NULL
#>
#> $value
#> NULL
#>
#> $method
#> [1] "MuZero recurrent inference (Schrittwieser et al. 2020)"
#>
#> attr(,"class")
#> [1] "morie_rich_result" "list"
