Speedup algorithm: Modified policy iterations

We approximate the continuous random model without labor:

by switching model with infinite many states in order:

.

.

.

where

.

.

.

for many states as in 0908

We calculate each policy function for each as it is set like:

independently for each i and calculate the next index number.

The random process can be

for as before.

When we compute each policy function for each, we apply modified policy iterations.

We also implement both error bounds and modified policy iterations.

Technically, we attach option 4 by modifying option 3 and add option 5 to option 1.