Information Technology Reference
In-Depth Information
in year 4000 BC, with fifty gold pieces, zero warriors, and one worker who collects
resources from a nearby tile. A city is either producing a warrior or collecting gold
pieces on any given turn. Alice can win by collecting 185 pieces of gold, and the
other agent can win by capturing Alice's city.
An experiment consists of running Alice against its opponent and noting the
results. Then, if Alice loses against the other agent, the human designer can exper-
iment with Alice's design by asking REM to adapt the model of the Alice agent to
avoid losing again. REM uses the results of the simulation to adapt Alice's agent
mode. The designer now evaluates the new model by running the simulation again.
The designer can repeat this Generate and Test cycle and experiment with Alice's
design until an acceptable design is produced.
17.4.5.1 Experiment #1
The purpose of the first experiment was to test whether REM could make a simple
adaptation to improve Alice's performance versus Freeciv's built-in robot player.
In general, a player of this reduced game has to make a decision about allocating
resources between collecting gold and creating warriors to defend its city. In the
first experiment, Alice's strategy was such that she devoted all of her resources to
the former pursuit. An obvious adaptation is to adjust Alice to balance her resource
allocation, and the first experiment tested whether REM could make this adaptation.
In the experiment, Alice played against Freeciv's robot player, which we call
Frank
, configured at its highest skill level. Although Alice had knowledge that Frank
could win by capturing her city, she was unaware that Frank had more powerful
weaponry and more production capacity than she had.
When played against Frank, Alice initially succumbed to his attacking chariots,
legions, and horsemen. Before losing, Alice was able to acquire 175 units of gold
and lived for 3075 years. However, Alice failed to accomplish her Goal of collecting
185 units of gold, thereby requiring retrospective adaptation.
In this experiment, no transformation was needed. That is, the failure was that an
Organizer rather than a Goal was flawed. Realizing a replacement Organizer took
place by inserting a new State, whose success would satisfy the preconditions of a
problem State. Such States are called
patch states
.
A patch State was created by first, searching a small library of generic Goal pat-
terns, each of which was checked to see if its success would satisfy the preconditions
of the problem State. After an instantiated Goal pattern was found it was assigned
as the Goal of the patch State. This patch State was then inserted into the local-
ized Organizer just prior to the problem State. This guarantees the problem State's
precondition is satisfied upon its visitation.
In Experiment #1, the patch State, was added with a Goal to build additional
warriors. This Goal increases the defense of Alice's city if she is visibly outgunned
on the game map.
After performing this adaptation, the new agent, Alice', was tested against Frank.
While still outgunned, Alice' fared better in longevity and defense. She lasted 3125
Search WWH ::
Custom Search