p. 870, l. 36
However, in practice one can get better results using higher-order ODE solvers, such as Heung’s
method
should be Heun's method, the second order ODE solver ?
and while I am here:
p. 1122, l34 This is a linear system of equations in |S| unknowns, We c −1 Alternatively, we can use value iteration by co
either . to end the sentence or a small we
and (sorry for having separate things here) next line:
Alternatively, we can use value iteration by computing...
but since in value iteration the policy implicitly gets updated, one either writes
Alternatively, we can iteratively solve by computing... or mention to use instead of the max_a the a from pi in the value iteration.
p. 870, l. 36 However, in practice one can get better results using higher-order ODE solvers, such as Heung’s method
should be Heun's method, the second order ODE solver ?
and while I am here: p. 1122, l34 This is a linear system of equations in |S| unknowns, We c −1 Alternatively, we can use value iteration by co either . to end the sentence or a small we
and (sorry for having separate things here) next line: Alternatively, we can use value iteration by computing... but since in value iteration the policy implicitly gets updated, one either writes Alternatively, we can iteratively solve by computing... or mention to use instead of the max_a the a from pi in the value iteration.