[x] Change section titles: A queueing model for the ED-EMS interface, Strategic manipulation of the ED-EMS interface
[x] Game subsections: 4.1 The model, 4.2 Results
Backwards induction becomes a sentence: ``We do this using backwards induction {throw couple of references} by first solving the game from the ambulance's point of view then from the hospital's point of view. One solution concept is N.E. {that becomes a paragraph and then another paragraph is...} another to be able to model the emergent behaviours is the learning algorithms and specifically the asymmetric replicator dynamics.
The way it is solved is first numerically for all pair of strategies we use the Brent's algorithm to find the BR of the ambulance center and that allow us to populate the utilities of all the hospitals because by solving the play of the EMS we know the arrival rate of every hospital which gives us this game.
1 line for backwards induction + 1 paragraph for N.E. + all equations for learning algorithm (maybe no change)
[x] Major contributions - queue + performance measures + game + emergent behaviour
In literature review should address what other papers lack and it's covered here
[x] Type 1+2 individuals definitions
Type 2 individuals are patients arriving in ambulances who can be blocked (usually patients that are deemed not to be critical)
Define on queueing section + point at the definition on game section
[x] Arriving state in blocking time:
Fix typo
Remove the word model (-> system)
[x] $C_i$ use number of servers -> at the start say that servers can be either healthcare professionals OR beds
[x] EMS -> Ambulance service, ED -> hospital
[ ] Make payoff matrices on results section more readable
RL algorithm (optional)
[ ] Plot results
[ ] Plot a 3rd graph with the impact of the server's behaviour on the hospital (waiting/blocking time)
Future things to keep in mind:
Talk for Greggynog, Vancouver, Finland
When service rate at a state is changed, change the surrounding states
What has been done:
To be discussed:
To be done: